Visualisation of smoothed particle hydrodynamics data using splash¶

Introduction¶

While many wonderful commercial software packages exist for visualising scientific data (such as the widely used Interactive Data Language), I found they were cumbersome for particle-based data. Much of what I wanted to do was specific to SPH, like interpolating to an array of pixels using the kernel. While generic routines exist for such tasks, I could not explain how they worked, and they were slow. Also, while interactive gizmos are handy, it was more difficult to perform the same tasks non-interactively, as required for the production of animations. The major work in the visualisation of SPH data is not the image production itself but the manipulation of data prior to plotting. Much of this manipulation makes sense within an SPH framework.

splash is designed for this specific task - to use SPH tools to analyse SPH data. Publishable images and animations can be obtained efficiently from the raw data with a minimum amount of user effort. The development of powerful visualisation tools has enabled me to pick up on effects present in my simulation results that I would not otherwise have noticed — the difference between a raw particle plot and a rendered image can be substantial. A key goal of splash is to eliminate the use of crap-looking particle plots as a means of representing SPH data.

What it does¶

splash is a utility for visualisation of output from (astrophysical) simulations using the Smoothed Particle Hydrodynamics (SPH) method in one, two and three dimensions. It is written in modern Fortran and utilises giza, a custom-build backend graphics library to do the actual plotting. The following features are included:

Rendered plots of particle data to an array of pixels using the SPH kernel
Cross section slices through 2D and 3D data (as both particle plots and rendered images).
Fast projections through 3D data (i.e., column density plots, or integration of other quantities along the line of sight)
Surface renderings of 3D data.
Vector plots of the velocity (and other vector quantities), including vector plots in a Cross section slice in 3D.
Rotation and animation sequence generation for 3D data.
Automatic stepping through timesteps, making animations simple to produce.
Interactive mode for detailed examination of timestep data (e.g. zooming, rotating, stepping forwards/backwards, log axes, adapting limits).
Remote visualisation via simple X-Windows forwarding
Multiple plots on page, including option to automatically tile plots if $y-$ and $x-$ limits are the same.
Plot limits can be fixed, adaptive or particle tracking.
Exact solutions for common SPH test problems (e.g. shock tubes, sedov blast wave).
Calculation of quantities not dumped (e.g. pressure, entropy)
Conversion of binary data files to ascii
Interpolation of SPH data to 2D and 3D grids
Plotting in different coordinate systems (e.g. cylindrical coordinates), for both coordinates and vector components
Rescaling of data into physical units.
Straightforward production of both bitmap (png) and vector (eps, pdf) images which can then be converted into animations or inserted into LaTeX documents.

Examples of plots produced with splash can be found in the gallery or by looking at papers with figures produced using splash.

What it doesn’t do¶

splash is geared towards gas dynamics simulations with SPH and has basically grown out of my visualisation needs. Thus it is not particularly useful for things like water and solids in SPH. An SPH visualisation tool geared towards the non-gaseous side of things you may want to have a look at is pv-meshless, by John Biddiscombe.

splash also doesn’t make coffee.

splash, the paper¶

The algorithms implemented in splash are not described here, but instead described in a paper [Price07], available from:

http://www.publish.csiro.au/?paper=AS07022

This paper should be cited if you use splash for scientific purposes, and please do so as it is my only form of thanks.

Version History¶

3.3.0: (20/05/21)

bug fix with surface density plot with physical units on
splash calc lightcurve computes spectra from local blackbody emission if T and kappa given
lightcurve now performs frequency-dependent ray tracing
added “–anglex”,”–angley”,”–anglez” flags
can add labelled arrows by typing ^ in interactive mode, also delete/edit
capital M, 0 or ncols+1 from main menu gives multiplot
added -multi flag for multiplot from command line

3.2.1: (26/04/21)

added –xsec=1.0 and –kappa=1.0 flags to specify cross section position and opacity, respectively
specifying –xsec automatically switches from projection to cross section
specifying –kappa turns on opacity rendering
bug fix in splash calc tracks
can use –track=1,2,3 to specify list of particles

3.2.0: (20/04/21)

disable ALL prompts if any command line flags set
all environment variables can now be given as command line flags using lower case string after last underscore e.g. SPLASH_CENTRE_ON_SINK=1 becomes –sink=1 on command line
useful options include –corotate, –sink=1, –debug and more
splash to grid recognises flags including –periodic, –npix=100,100,100 and –convert=1,4
added -gandalf and -f gandalf as shortcut for seren data read
assume default xw device and disable device prompt if any command line flags set
s/S options now do the same thing

3.1.1: (31/03/21)

automatically plot y vs x given a two-column data file
planet wake coordinate system added
bug fix with SPLASH_COROTATE
bug fix reading phantom dumps when number of particles of each type does not match itype array
bug fixes in grid2pdf

3.1.0: (16/02/21)

splash calc lightcurve implemented
sink particles ON by default
changing units rescales plot limits correctly
further improvements to ray tracing / opacity rendering with physical opacity
can change units temporarily without writing .units file
auto-select closest velocity and mass unit and better default time unit in phantom/sphNG read
error message if Inf or NaN read from .units file
bug fix with units prompt
floating colour bars are white not black
automatically write copyright in Hollywood mode
auto-render fits files
read softening length from phantom sinks if accretion radius is zero

3.0.2: (20/01/21)

opacity rendering uses physical value of kappa, can also use opacity defined on particles
can track multiple particles with ‘splash calc tracks’ by specifying ids in splash.tracks file
support for SWIFT code in gadget_hdf5
auto-recognise format for .csv files
improved starsmasher data read
improved physical unit selection
exact solution lines can be plotted in background colour
bug fix for dead particles in phantom dumps
seg fault in fits reader fixed
seg fault in gadget data read fixed
bug fix in x-menu options

3.0.0: (26/08/20)

Unified splash binary with -f flag to specify format
automated format recognition for phantom, gadget (and hdf5 variants) and fits
cleaner d) menu
splash is compiled in double precision by default
rotation settings used in splash to grid to rotate particles
bug fix in mbatesph data read
pysplash utility for reading SPH data formats into python
libsplash.so, libexact.so and libread.so libraries

2.10.1: (24/06/20)

exact solution can appear in legend
can also plot under data
fits reader and denoise utility can read/write spectral cubes
text shapes can print header variables using %(var)
can shift cross section by precise amounts in interactive mode using number followed by u/d
fits reader includes header quantities
reduced verbosity for non-interactive plots
use of fake dust particles is now via menu option, not environment variable
max particle types = 24
userguide in readthedocs format
bug fix with save limits with particle tracking
support for .pfm pixelmap format as output
physical units are ON by default
prompts only for particle types present in data

2.10.0: (14/02/20)

much improved splash to grid - bug fixes with pixel number and roundoff error
use Petkova (2018) method for sub-pixel rendering to 3D grid and 3D projections
added bytestream output formats for splash to grid and splash to ascii
can press number and -/+ to zoom out/in by that factor in interactive mode
use SPLASH_COROTATE=1,3 to corotate with arbitrary pair of sink particles
SPLASH_COROTATE also gives velocity field in corotating frame
splash to ascii can write particular columns by setting SPLASH_CONVERT=1,4
plasma beta correct in both code and physical units
working fits reader and splash-denoise utility

2.9.1: (08/11/19)

cleaner menu options for units and calculated quantities
surface rendering allowed with 3D perspective turned off
automatic labelling of grain sizes in density and column density plots
adaptive limits on log colour bars show 3 dex range by default
auto-adjust limits to device aspect ratio works with multiple panels
bug fixes with r-z rendering
Toomre Q prompts for mass

2.9.0: (05/04/19)

general header quantities are read and available in function parser
more robust label detection and parsing during ascii data read
splash to grid works in non-cartesian geometries
added flared and log-flared coordinate systems
Doppler shift colour bar
can customise line style and colour when plotting multiple exact solutions
seg faults fixed
better plot tiling decisions
disappearing arrows bug fix
Rafikov disc- planet exact solution added
atan2 implemented in function parser
various multigrain phantom read fixes (incl. seg faults)
exact rendering implemented in 2D
libsplash implemented for use as Python splash backend

2.8.0: (06/04/18)

360/4pi video mode added
automatically read labels from ascii file headers
nearest sensible unit (e.g. au or pc) used by default
cactus hdf5 data read
kernel-smoothed particle plots of arbitrary quantities
Viridis, Ocean and Inferno colour schemes
can customise line colours
Bondi flow exact solution
option for ticks but no labels
correct units in surface density plots
colour bar on top or left
support for multi-grain dust in Phantom
bug fix with NaNs in ascii files

2.7.0: (03/05/17)

Hollywood mode added (ctrl-m in interactive mode)
better handling of dust/gas phantom data
added rotated cartesian geometry
rendering implemented in r-phi coordinates
added Fortran 2008 intrinsics to function parser
better rectangle plotting
better falcON data read
Ogilvie-Lubow exact solution for planet-disc interaction
tipsy read now works when splash compiled in double precision
splash to gridascii2 implemented
bugs with r-phi rendering fixed

2.6.0: (22/10/15)

SILO, falcON and .pbob data reads implemented
bug fixes in gadget-hdf5 reader
can recognise particle types in ascii read
more robust sphNG read
dust fraction recognised in phantom data read
Toomre Q works in physical units
bug fix with disappearing units labels
bug fix in shock tube exact solution
added splash calc delta
splash to ascii keeps precision
better power spectra

2.5.1: (29/01/15)

error bar style options
support for 5K displays
can plot vectors and render with colours if h not read
range restrictions apply during splash to grid
improved line-style legend
now up to 6 line styles
fixes to amuse-hdf5 read
phantom read handles star/dm particles
various bugs fixed

2.5.0: (22/08/14)

instant multiplots by giving multiple columns as y axis
ability to plot multiple exact solution files on same plot
compiles in parallel by default
support for tagged sphNG/Phantom format
AMUSE hdf5 format reader added
various bug fixes

2.4.1: (01/04/14)

Roche-lobe plotting vastly improved
newunit= issue fixed
bug fix with reading sink velocities from Phantom
other minor bug fixes.

2.4.0: (21/02/14)

time formatting in legend can include general functions like %(t + 1000)
option to include sinks in opacity rendering
supports one-fluid dust visualisation
C-shock exact solution
better polytrope solution

2.3.1: (11/11/13)

SPLASH_COROTATE option to plot in frame corotating with sinks
bug fixes with handling of dead/accreted/boundary particles in sphNG/phantom
various other bugs fixed.

2.3.0: (09/08/13)

can customise time formatting in legend
improvements to legends
less verboseness
splash can read and plot pixel maps produced with -o ascii
3D vector field plotting improved
bug fix with gfortran 4.8

2.2.2: (10/05/13)

particle tracking by type implemented
can interpolate specific columns in splash to grid
SPLASH_CENTRE_ON_SINK option generic to all data reads
Aly Reheam format added
option for 2nd y axis on plots
bug fix with X11 linking on Ubuntu
can read gadget ICs files

2.2.1: (21/02/13)

minor bug with axes plotting fixed
Wendland kernels added
bugs with exact solution plotting fixed
bug fix with tracking of dark matter particles

2.2.0: (16/11/12)

option to use different kernels for interpolation
floating/inset colour bars added
splash to gadget conversion implemented
splash to grid works in 2D
improved interfaces to shapes and animation sequences
automatically turns on dark matter particle plotting if no gas
interactive mode help displayed automatically

2.1.1: (31/08/12)

irregular/circular particle selection using shift-left/middle click
improved h5part and GADGET HDF5 data reads
splash can be compiled in double precision
bug fixes with calculated quantities + change of coordinate systems
improved vector plot legend
option for box+numbers but no labels added

2.1.0: (16/05/12)

3D vector field visualisation added
GADGET HDF5 read implemented
page sizes can be specified in pixels
limits can auto-adapt to device aspect ratio
more general exact solution from file option
tiling works with one colour bar per row
splash calc handles different particle types

2.0.0: (29/08/11)

new giza backend - antialiased lines
real fonts
pdf, eps and svg drivers
fewer build dependencies (only cairo, X11)
support for semi-transparent text
Double rendering (with transparent background) implemented.

1.15.0: (29/08/11)

Multiplot with different particle types implemented
calculated quantities list is now pre-filled automatically
preliminary support for r-phi and r-z rendering
outlined solid markers implemented
better handling of multiple types
manual contour levels can be specified in splash.contours
parallel splash to grid
better support for non-square pixels
clipping of numbers at edge of viewport fixed

1.14.1: (17/03/11)

SEREN data read added
dragon read updated
build follows Gnu conventions on DEST and DESTDIR (needed for macports build)
can have up to 12 particle types
exact solutions re-ordered
dusty wave exact solution added

1.14.0: (06/12/10)

Can flip between rendered quantities in interactive mode using ‘f/F’
SPLASH_DEFAULTS variable can be set for system-wide defaults
can plot arbitrary functions of x,t as exact solution
asplash better handles blank lines in header and can specify time, gamma location with env. variables
added data read for the H5PART format
GADGET read across multiple files implemented
VINE read works with particle injection
error bars can be plotted for both x and y axis simultaneously
default rotation angles are set if 3D perspective turned on
new directory layout and more helpful error messages during build
PGPLOT linking is easier to get right.

1.13.1: (26/02/10)

bugs with new calc_quantities module fixed
generic library interface implemented so backend can be changed easily
bug fix with auto pixel selection
simpler foreground/background colour setting
added subgrid interpolation warning

1.13.0: (25/02/10)

function parser incorporated
calculated quantities can now be specified at runtime, arbitrary function plotting implemented as an exact solution
command-line SPH->grid conversion (“splash to grid”) implemented
ctrl-t in interactive mode adds arbitrary text box
better line style/colour changing
bug fix with tiling and y-axis labels
various other bug fixes.

1.12.2: (15/07/09)

Variable marker sizes added, can plot particles as circles with size proportional to h
dark matter rendering with block-labelled GADGET format fixed
VINE read handles star particles
TIPSY read with ifort10.0.0 works
snsph read added
splash to phantom added
does not override labels for coords, vectors by default
bug fixes with contouring options
stability bug fixes with older compilers
more robust memory handling
bug fix with automatic pixel selection causing seg fault.

1.12.1: (20/04/09)

Can edit/delete text shapes interactively, also the colour bar label
can customise the label on projection plots
contour levels better defined
SPLASH_HMIN_CODEUNITS added
option for numeric labelling of contours
contour limits can be set separately to render limits for same quantity
minor bug fixes.

1.12.0: (22/12/08)

Command-line plotting implemented
ln transform added
bug fixes in GADGET read
Backspace over annotation (legends,titles,axes,colour bar) in interactive mode removes it
“splash calc” command line utility calculates time sequences of global quantities from a sequence of dump files
bug fix causing seg fault.

1.11.1: (13/10/08)

automatic number of pixels and exact pixel boundaries implemented
mass does not have to be read from dump file
frame changes are per-page not per-dump file for animation sequences
lower stacksize footprint
bug fix with circles of interaction
bug fixes with block-labelled GADGET read
Steve Foulkes data read added.

1.11.0: (15/08/08)

ability to use subset of particles in restricted parameter range(s)
probability density function plot option
plot-hugging colour bars added
ability to annotate plot with a range of shapes
v,V,w and H implemented in interactive mode for >1 panel
various bug fixes (including one with vphi).

1.10.2: (08/05/08)

disc surface density / toomre q parameter plotting added
flash colour schemes added
splash to binary convert option
can change order in which particle types are plotted
splash.columns file overrides column label settings
vanaverbeke format read
various bug fixes.

1.10.1: (11/03/08)

“splash to” command line option converts binary dumps to ascii format
vector plots + rotation now implemented
block labelled GADGET format read
ring-spreading exact solution added.

1.10.0: (28/11/07)

horizontal colour bars implemented
-p, -o command line options
can have mixed types in data reads
TIPSY and DRAGON data reads
density weighted rendering
normalisation applies to column density plots
improved particle tracking
save as option
various bug fixes

1.9.2: (12/09/07)

improvements to ascii read including asplash -e option
smarter foreground/background colour changing for titles
min=max problem fixed (caught by splash not pgplot)
fixed vector arrow length option
other minor changes and bug fixes

1.9.1: (11/07/07)

environment variables + improvements to gadget data read
better prompting
3 new colour schemes
improved legend/title options
other minor changes

1.9.0: (21/05/07)

animation sequences implemented
origin settings now affect radius calculation and are relative to tracked particle
automatic line width choice for postscript devices
w key adapts vector arrows
vastly improved userguide

1.8.1: (28/03/07)

option to hide vector arrows where there are no particles added
smoother 3D plotting at low pixel numbers
smoother vector plots
bug fixes with a)
issues with round-off error with z integration of vectors fixed.

1.8.0: (14/03/07)

hidden particles not used in rendering
units for z integration added
1. & g) implemented in interactive mode for multiple-plots-per-page
improved cross section using x in interactive mode

1.7.2: (19/02/07)

Menu shortcuts implemented
bug fix/ more sensible transformation of angular vector components in different co-ordinate systems
improvements to interactive zoom and origin recentreing
improved colour-by-type option
restrictions on page size removed
minor bug fixes

1.7.1: (04/01/07)

command line options for defaults and limits files added
minor bug fixes

1.7.0: (13/12/06)

renamed SPLASH instead of SUPERSPHPLOT
much faster data read for gadget and sphNG reads (only required columns read)
physical units can be saved to file
new menu formats
various other bug fixes

1.6.2: (24/10/06)

fast particle plotting and streamline plotting implemented
more bug fixes with interactive mode on multiplots
various other bug fixes

1.6.1: (24/8/06)

bug fixes to 1.6.0, further improvements to interactive mode on multiplots

1.6.0: (10/8/06)

Interactive mode on multiple plots per page
highly optimised interpolation + parallel version
new Makefile
various bug fixes

1.5.4: (06/7/06)

Handles multiple SPH/non-SPH particle types
axes redrawn after rendering
minor bug fixes

1.5.3: (27/6/06)

minor bug fixes/improvements to multiple plots per page
colour bar labelling tiled plots
legend
Accelerated rendering option for projections.

1.5.2: (11/5/06)

“S)” option for saving limits and defaults
MUCH faster interactive replotting (no unnecessary re-rendering)
a few other minor things

1.5.1: (26/4/06)

docs updated for v1.5, other minor changes

1.5.0: (17/3/06)

3D perspective added
3D opacity rendering
improved rotation, colour schemes
adjustable vector arrows (+legend)
improved timestepping behaviour
speed enhancements
physical unit rescaling

1.0.5: (28/9/05)

error calculation for exact solutions
legend for plot markers
exact_densityprofiles added
more colour schemes
unit rescaling improved
other minor changes + bug fixes

1.0.4: (17/8/05)

better colour schemes
interactive colour scheme changing
various minor changes and bug fixes

1.0.3: (5/7/05)

rescale data option
better page setup
improved zooming
interactive particle tracking
various minor changes and bug fixes

1.0.2 :

much improved ascii data read
better line plotting
zoom on powerspectrum plots + various bug fixes

1.0.1 :

bug fixes relating to colour bars on multiplots

1.0.0 :

first official release
version given to many people at IPAM meeting and put on web

License¶

splash - a visualisation tool for SPH data ©2004-2021 Daniel Price. This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.

Getting started¶

Installing from your package manager¶

Pre-packaged builds of splash exist for most operating systems.

Stable version¶

Mac OS via homebrew:

brew tap danieljprice/all
brew install splash

Mac OS via Macports:

sudo port install splash

Linux or Windows Linux Subsystem (Ubuntu):

sudo apt-get install splash

Development version¶

SPLASH and giza (the plotting backend) both have public repositories, so you can check out the latest and greatest code at any time. Both the splash and giza repositories are generally stable, so it is usually safe to get the very latest version. Just use:

Mac OS via homebrew:

brew tap danieljprice/all
brew install --HEAD splash

or compile from source following the instructions below.

Compiling from source¶

Basic instructions¶

If you have admin (super user) permissions:

git clone https://github.com/danieljprice/giza.git; cd giza; ./configure; make; sudo make install; cd ..
git clone https://github.com/danieljprice/splash.git
cd splash; make SYSTEM=gfortran; sudo make install

Installing in your home space¶

If you do not have admin permissions. That is, to install in your home space:

git clone https://github.com/danieljprice/splash.git
cd splash; git clone https://github.com/danieljprice/giza.git
make SYSTEM=gfortran withgiza

Important

If you have installed splash in your home space, you will need to set the following environment variables for everything to work. Put the following commands in your ~/.bashrc file or equivalent, so they are set every time you log in:

export SPLASH_DIR=$HOME/splash
export PATH=$PATH:$SPLASH_DIR/bin
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$SPLASH_DIR/giza/lib

Advanced installation guide¶

The basic steps for installation are as follows:

make sure you have a recent Fortran compiler (such as gfortran)
make sure you have cairo on your system
compile splash and giza

Fortran compilers¶

Numerous Fortran 90/95/2003 compilers exist. The most widely available are:

gfortran, the free Gnu Fortran Compiler http://gcc.gnu.org/wiki/GFortran
ifort, one of the most widely available commercial compilers (and is very good) with (limited) free licence for Linux. http://software.intel.com/en-us/articles/intel-compilers/

Both of these successfully compile splash and the giza library.

Cairo graphics library¶

Cairo is a low-level system library used in many applications. Thus it is highly likely that you already have a copy on your system and already in your library path. Look for the header file cairo.h, e.g. using:

locate cairo.h

or have a look in the usual places (e.g. /usr/include/cairo, /usr/X11/include). If not, you can usually use your inbuilt package manager to install cairo as follows:

Debian/Ubuntu:
sudo apt-get install libcairo2-dev

Fedora/Red Hat/CentOS:
sudo yum install cairo-devel

OpenSUSE:
zypper install cairo-devel

MacPorts:
sudo port install cairo

Alternatively, use the script provided in the root-level splash directory:

./install-cairo.sh

which downloads and installs both pixman and cairo into the giza/ subdirectory. Unlike the methods above, this does not require any admin/superuser permissions.

Compiling and linking with giza¶

You can either install giza with your package manager, or in a subdirectory of splash. To install in a splash subdirectory, use:

     cd splash
     git clone http://github.com/danieljprice/giza
make withgiza

of splash.

With giza installed via your package manager (or previously compiled as below), use:

cd splash
make GIZA_DIR=/usr/local

where GIZA_DIR points to the directory where giza was installed. To install giza in a splash subdirectory, use:

cd splash
git clone http://github.com/danieljprice/giza
make withgiza

A successful make will produce a binary called splash

Reading your data¶

The most important part is getting splash to read *your* data format. If you are using a publicly available code, it is reasonably likely that I have already written a read data subroutine which will read your dumps. If not it is best to look at some of the other examples and change the necessary parts to suit your data files. Note that reading directly from unformatted data files is *much* faster than reading from formatted (ascii) output.

A standard make will create a binary which supports the file formats listed in Binaries and data reads compiled by default, plus a bunch of others (type splash --formats to see what formats your build supports). All data formats in the splash repository that do not have an additional dependencies (e.g. HDF5) will be supported in the splash binary as of version 3.0.0. This means that the user needs to specify the data type they are reading as a command line option. For example, the following will read a phantom dumpfile:

splash --format phantom disc_00000

In some cases, the format of the file can be inferred if the the file has a known suffix. For example, the above line can be changed if the suffixe of the file is recognised:

splash disc_00000.pb

This will automatically recognise a Phantom binary dumpfile. For backwards compatibility with previous version of splash, one can add aliases into their .bashrc, or equivalent:

alias asplash='splash ' # Alias for ascii splash
alias ssplash='splash -f phantom '
alias gsplash='splash -f gadget '
alias vsplash='splash -f vine '
alias nsplash='splash -f ndspmhd '
alias rsplash='splash -f srosph '
alias dsplash='splash -f dragon '
alias srsplash='splash -f seren '
alias tsplash='splash -f tipsy '
alias tsplash='splash -f tipsy '
alias msplash='splash -f mhutch '

If splash is compiled with HDF5=yes, the formats listed in Supported HDF5 data formats will also be available in the splash binary. Other supported formats are listed in Other supported file formats that require external libraries, but these require additional libraries.

Binaries and data reads compiled by default¶
`splash` command	Format Read	`read_data` File	Comments
`splash -gadget <file>`	ascii	`read_data_asci.f90`	Generic data read for n-column ascii formats. Automatically determines number of columns and skips header lines. Can recognise SPH particle data based on the column labels. Use `splash -e` to plot non-SPH data (e.g. energy vs time files).
`splash -dragon <file>`	dragon	`read_data_dragon`	See environment variable options.
`splash -gadget <file>`	gadget, gadget-2, gadget-3	`read_data_gadget.f90`	Handles both default and block-labelled formats (see environment variable options).
`splash -ndspmhd <file>`	ndspmhd	`read_data_ndspmhd.f90`	Format for the ndspmhd SPH/SPMHD code (publicly available from my website).
`splash -phantom <file>`	sphNG, Phantom	`read_data_sphNG.f90`	sphNG is Matthew Bate’s SPH code. Option `-sphng` also works.
`splash -magma <file>`	magma	`read_data_srosph.f90`	Stephan Rosswog’s code
`splash -seren <file>`	seren	`read_data_seren.f90`	The SEREN SPH code (Hubber, McLeod et al.)
`splash -gasoline <file>`	gasoline, tipsy	`read_data_tipsy.f90`	Reads both binary and ascii TIPSY files (determined automatically). Option `-tipsy` also works.
`splash -vine <file>`	vine	`read_data_fine.f90`	See environment variable options.
`splash -starsmasher <file>`	StarSmasher	`read_data_starsmasher.f90`	The StarSmasher code (Gaburov et al. 2018)

Below is a list of the supported data formats that require HDF5.

Supported HDF5 data formats¶
`splash` Command	Read Format	`read_data` File	Comments
`splash -gadget_hdf5 <file>`	gadget HDF5 Files.	`read_data_gadget_hdf5.f90`	Reads HDF5 format from the gadget code.
`splash -amuse <file>`	AMUSE HDF5	`read_data_amuse_hdf5.f90`	Reads HDF5 format from the AMUSE framework.
`splash -cactus_hdf5 <file>`	Cactus HDF5	`read_data_cactus_hdf5.f90`
`splash -flash_hdf5 <file>`	FLASH tracer particles	`read_dataflash_hdf5.f90`	Reads tracer particle output from the FLASH code. The option `-flash` will also work.
`splash -falcon_hdf5 <file>`	falcON	`read_data_falcON_hdf5.f90`	Walter Dehnen’s SPH code format. The option `-falcon` will also work.

If the HDF5 read files end with .h5, the suffix _hdf5 from the splash command can be removed. For example:

splash -gadget dump_000.h5

will recognise that the file dump_000.h5 is in the HDF5 format, and will automatically select the correct read_data routine.

Below is a list of other formats supported, but have additional library requirements.

Other supported file formats that require external libraries¶
`splash` Command	Read Format	`read_data` File	Comments
`splash -pbob <file>`	PBOB Files	`read_data_pbob.f90`	Requires the PBOB Library. Compile `splash` with `PBOB_DIR=/path/to/pbob/`.
`splash -h5part <file>`	H5Part Files	`read_data_h5part.f90`	Requires the H5Part and HDF5 libraries. Compile `splash` with `H5PART_DIR=/path/to/h5part/`.
`splash -fits <file>`	FITS files	`read_data_fits.f90`	Requires FITS libraries. Try to compile `splash` with `FITS=yes`. If this does not work, point to the location of your fits libraries with `FITS_DIR=/path/to/fits`.

Further details on writing your own subroutine are given in appendix Data reads and command line options. The *easiest* way is to i) email me a sample data file and ii) the subroutine you used to write it, and I will happily create a data read for your file format.

Command line options¶

Typing splash --help gives a complete and up-to-date list of options. Currently these are:

Command line options:

 -p fileprefix     : change prefix to ALL settings files read/written by splash
 -d defaultsfile   : change name of defaults file read/written by splash
 -l limitsfile     : change name of limits file read/written by splash
 -e, -ev           : use default options best suited to ascii evolution files (ie. energy vs time)
 -lm, -lowmem      : use low memory mode [applies only to sphNG data read at present]
 -o pixformat      : dump pixel map in specified format (use just -o for list of formats)
 -f                : input file format to be read (ascii is default)

To select data formats, use the shortcuts below, or use the -f or --format command line options
Multiple data formats are not support in a single instance.
Supported data formats:
 -ascii            : ascii file format (default)
 -phantom -sphng   : Phantom and sphNG codes
 -ndspmhd          : ndsphmd code
 -gadget           : Gadget code
 -seren            : Seren code
..plus many others. Type --formats for a full list

The following formats support HDF5:
 -flash            : FLASH code
 -gadget           : Gadget code
 -cactus           : Cactus SPH code
 -falcon           : FalcON code
 -amuse            : AMUSE Framework

add a suffix "_hdf5" to the above command line options if your data files do not end with .h5.

Command line plotting mode:

 -x column         : specify x plot on command line (ie. do not prompt for x)
 -y column         : specify y plot on command line (ie. do not prompt for y)
 -r[ender] column  : specify rendered quantity on command line (ie. no render prompt)
                     (will take columns 1 and 2 as x and y if -x and/or -y not specified)
 -vec[tor] column  : specify vector plot quantity on command line (ie. no vector prompt)
 -c[ontour] column : specify contoured quantity on command line (ie. no contour prompt)
 -dev device       : specify plotting device on command line (ie. do not prompt)

 convert mode ("splash to X dumpfiles"):
 splash to ascii   : convert SPH data to ascii file dumpfile.ascii

        to binary  : convert SPH data to simple unformatted binary dumpfile.binary
                      write(1) time,npart,ncolumns
                      do i=1,npart
                         write(1) dat(1:ncolumns),itype
                      enddo
        to phantom : convert SPH data to binary dump file for PHANTOM
        to gadget  : convert SPH data to default GADGET snapshot file format

 Grid conversion mode ("splash to X dumpfiles"):
    splash to grid         : interpolate basic SPH data (density, plus velocity if present in data)
                             to 2D or 3D grid, write grid data to file (using default output=ascii)
           to gridascii    : as above, grid data written in ascii format
           to gridbinary   : as above, grid data in simple unformatted binary format:
                                write(unit) nx,ny,nz,ncolumns,time                 [ 4 bytes each ]
                                write(unit) (((rho(i,j,k),i=1,nx),j=1,ny),k=1,nz)  [ 4 bytes each ]
                                write(unit) (((vx(i,j,k), i=1,nx),j=1,ny),k=1,nz)  [ 4 bytes each ]
                                write(unit) (((vy(i,j,k), i=1,nx),j=1,ny),k=1,nz)  [ 4 bytes each ]
                                write(unit) (((...(i,j,k),i=1,nx),j=1,ny),k=1,nz)  [ 4 bytes each ]
        allto grid         : as above, interpolating *all* columns to the grid (and output file)
        allto gridascii    : as above, with ascii output
        allto gridbinary   : as above, with binary output

 Analysis mode ("splash calc X dumpfiles") on a sequence of dump files:
  splash calc energies     : calculate KE,PE,total energy vs time
                             output to file called 'energy.out'
         calc massaboverho : mass above a series of density thresholds vs time
                             output to file called 'massaboverho.out'
         calc max          : maximum of each column vs. time
                             output to file called 'maxvals.out'
         calc min          : minimum of each column vs. time
                             output to file called 'minvals.out'
         calc diff           : (max - min) of each column vs. time
                             output to file called 'diffvals.out'
         calc amp          : 0.5*(max - min) of each column vs. time
                             output to file called 'ampvals.out'
         calc delta        : 0.5*(max - min)/mean of each column vs. time
                             output to file called 'deltavals.out'
         calc mean         : mean of each column vs. time
                             output to file called 'meanvals.out'
         calc rms          : (mass weighted) root mean square of each column vs. time
                             output to file called 'rmsvals.out'

  the above options all produce a small ascii file with one row per input file.
  the following option produces a file equivalent in size to one input file (in ascii format):

         calc timeaverage  : time average of *all* entries for every particle
                             output to file called 'time_average.out'

         calc ratio        : ratio of *all* entries in each file compared to first
                             output to file called 'ratio.out'

Command-line options can be entered in any order on the command line (even after the dump file names). For more information on the convert utility (splash to ascii) see Conversion of binary data files to ascii. For details of the -o ppm or -o ascii option see Outputting the raw pixel map to a file. For details of the -ev option, see Using splash to plot energy vs time plots.

Options affecting all data reads¶

Command line flags (or environment variables) that affect all data reads are:

—defaults	SPLASH_DEFAULTS	gives name of system-wide `splash.defaults` file (and splash.limits etc.) that will be used if there is none in the current dir. e.g. `--defaults=$HOME/splash.defaults`
—kernel	SPLASH_KERNEL	changes the smoothing kernel used in the interpolations (e.g. `cubic` or `quintic`). Can also be changed in the (r)endering options.
—debug	SPLASH_DEBUG	if set to `yes` or `true`, turns on verbose debugging output. Useful to trace code crashes (but of course, this never happens…).
—sink=1	SPLASH_CENTRE_ON_SINK	if set to a number n, centres coordinates and velocities on the nth sink/star particle (e.g. `export SPLASH_CENTRE_ON_SINK=2`)
—corotate	SPLASH_COROTATE	plot in corotating frame based on locations of 2 sink particles (e.g. `--corotate=1,3`)
—beam=2.0	SPLASH_BEAM	if given a value $>$0 enforces a minimum smoothing length, specified in code units, all the particles. Useful to “dumb-down” the resolution of SPH simulations to match observational resolution. If this variable is set the “accelerated rendering” option in the (r)endering options is also turned on as otherwise slow rendering can result.
—xmin=0.1 —xmax=0.1 —ymin=0.1 —ymax=0.1	SPLASH_MARGIN_XMIN SPLASH_MARGIN_XMAX SPLASH_MARGIN_YMIN SPLASH_MARGIN_YMAX	can be used to manually adjust the left page page margin (set to fraction of viewport, negative values are allowed).

Ascii data read¶

For several data reads there are command-line flags which can be set at runtime which are specific to the data read. For the ascii data read (splash -f ascii) these are:

`--ncolumns=10`	if given a value $>$0 sets the number of columns to be read from ascii data (overrides the automatic number of columns determination).
`--nheaderlines=3`	if given a value $>=$0 sets the number of header lines to skip (overrides the automatic determination).
`--columnsfile=/home/me/mylabels`	can be used to provide the location of (path to) the default `columns` file containing the labels for ascii data. Overridden by the presence of a local `columns` file.
`--time=3.0`	if given a nonzero value sets the time to use in the legend (fixed for all files)
`--gamma=1.667`	if given a nonzero value sets gamma to use in exact solution calculations (fixed for all files)
`--timeheader=1`	sets the integer line number where the time appears in the header
`--gammaheader=3`	sets the integer line number where gamma appears in the header

The above options can be set as environment variables by prefixing them with ASPLASH, e.g.:

export ASPLASH_NCOLUMNS=10
splash datafile.txt

GADGET data read¶

For the GADGET read (splash -f gadget) the options are:

`--format=2`	if set = 2, reads the block labelled GADGET format instead of the default (non block labelled) format.
`--usez`	if `yes` or `true` uses the redshift in the legend instead of code time.
`--hsoft=500.`	if given a value $>$ 0.0 will assign a smoothing length to dark matter particles for which rendered plots of column density can then be made.
`--extracols`	if set to a comma separated list of column labels, will attempt to read additional columns containing gas particle properties beyond the end of the file (not applicable if –format=2).
`--starpartcols`	if set to a comma separated list of column labels, will attempt to read additional columns containing star particle properties beyond the end of the file (and after any extra gas particle columns) (not applicable if GSPLASH_FORMAT=2).
`--checkids`	if set to `yes` or `true`, reads and checks particle IDs, excluding particles with negative IDs as accreted (gives them a negative smoothing length which means they are ignored in renderings).
`--hcolumn`	if set to a positive integer, specifies the location of the smoothing length in the columns, overriding any default settings.
`--ignore-iflagcool`	if set,does not assume that extra columns are present even if the cooling flag is set in the header.

For backwards compatibility, the above options can also be set as environment variables by prefixing them with GSPLASH, e.g.:

export GSPLASH_FORMAT=2
splash -gadget snap_00010

which is equivalent to:

splash -f gadget --format=2 snap_00010

For the GADGET read gsplash will also look for, and read if present, files called snapshot_xxx.hsml and/or snapshot_xxx.dens (where snapshot_xxx is the name of the corresponding GADGET dump file) which contain smoothing lengths and/or a density estimate for dark matter particles (these should just be one-column ascii files).

VINE data read¶

For the VINE read (splash -vine) the options are:

`--hfac`	if `yes` or `true` multiplies smoothing length read from the dump file by a factor of 2.8 (for use with older VINE dumps where the smoothing length is defined as in a Plummer kernel rather than as the usual SPH smoothing length).
`--mhd`	if set, reads VINE dumps containing MHD arrays (note that setting VINE_MHD also works).

sphNG data read¶

For the sphNG and PHANTOM read (splash -phantom) the options are:

`--cm`	resets the positions such that the centre of mass is exactly at the origin.
`--omega=3.142`	if non-zero, subtracts solid body rotation with omega as specified to give velocities in co-rotating frame
`--omegat=3.142`	same as –omega but applies to velocities also
`--timeunit=hrs`	sets default time units, either ’s’, ’min’, ’hrs’, ’days’, ’yr’ or ’tfreefall’ (used verbatim in legend)

dragon data read¶

For the dragon read (splash -dragon) the options are:

--extracols

specifies number of extra columns present in the file which are dumped after the itype array

Stephan Rosswog data read¶

For the srosph read (splash) the options are:

`--format=MHD`	can be `MHD` or `HYDRO` which read the appropriate data format from either the MHD or hydrodynamic codes
`--com`	if set resets the positions such that the centre of mass is exactly at the origin.
`--corotating`	velocities are transformed to corotating frame
`--hfact=1.2`	can be changed to give correct parameter in $h=h_{fact}(m/\rho)^{1/3}$ used to set the particle masses when rendering minidumps (i.e., when the mass is not dumped). Default is RSPLASH_HFACT=1.5

ndspmhd data read¶

For the ndspmhd read (splash -ndspmhd) the options are:

--barycentric

plots barycentric quantities for one-fluid dust instead of creating fake second set of particles

H5Part data read¶

For the H5PART read (splash -f h5part) the options are:

`--ndim=3`	number of spatial dimensions $d$ (overrides value inferred from data)
`--hfac=1.2`	factor to use to compute h from $h = h_{fac} *(m/\rho)^{1/d}$ if h not present in data
`--hsml=3.0`	value for global smoothing length h (if h not present in data)
`--typeid=MatID`	name of the dataset containing the particle type identification (default is “MatID”)

Environment variables¶

Several runtime options for splash can be set using environment variables. These are variables set from your unix shell. In the bash shell, environment variables are set from the command line using

export VAR='blah'

or by putting this command in your .bash_profile/.bashrc. In csh, the equivalent is

setenv VAR 'blah'

or by putting the above in your .cshrc file.

Changing the font¶

Several environment variables affect the backend plotting library. Probably the most useful is the ability to change font:

export GIZA_FONT='Helvetica'

where the name is a reasonable guess as to the font you want to use (the default is Times). In particular, if you are having trouble displaying unicode characters such as greek letters, you can just change the font until you find one that works.

Endian changing¶

On some compilers, the endian-ness (byte order) when reading unformatted binary data files can be changed at runtime. This is useful for looking at files on different systems to the one on which they were created (e.g. x86 machines create little-endian files by default, whereas IBM/powerpc machines create big-endian). Environment variables for changing the endian-ness of the data read for some common compilers are given below:

Compiler	Environment variable	Big endian	Little endian	Other
gfortran	`GFORTRAN_CONVERT_UNIT`	`big_endian`	`little_endian`	`swap`
ifort	`F_UFMTENDIAN`	`big`	`little`

For compilers without this feature, almost all can change the endian-ness at compile time, and the appropriate flags for doing so can be set using

export ENDIAN='BIG'

or LITTLE before compiling splash (this adds the appropriate compile-time flags for the compiler selected using the SYSTEM environment variable in the splash Makefile).

Basic splash usage¶

Simple two column plot¶

Once you have successfully compiled splash with a read data file that will read your data format, splash is invoked with the name of the data file(s) on the command line, e.g.

splash myrun*.dat

where splash should be replaced with ssplash, gsplash etc. depending on the data format.

After a successful data read, the menu should appear as something like the following (the example given is for a “minidump” from Stephan Rosswog’s SPH code):

dprice$ rsplash minidump.00001

    _                                                 _
   (_)   _               _           _         _     (_)_
      _ (_)    ___ _ __ | | __ _ ___| |__     (_)   _  (_)
   _ (_)  _   / __| '_ \| |/ _` / __| '_ \       _ (_)
  (_)  _ (_)  \__ \ |_) | | (_| \__ \ | | |  _  (_) _
      (_)  _  |___/ .__/|_|\__,_|___/_| |_| (_)  _ (_)
          (_)  (_)|_| (_) (_)  (_)(_) (_)(_) (_)(_)

  ( B | y ) ( D | a | n | i | e | l ) ( P | r | i | c | e )

...etc...

 You may choose from a delectable sample of plots
-------------------------------------------------------
  1) x                     7) particle mass
  2) y                     8) B\dx
  3) z                     9) B\dy
  4) h                    10) B\dz
  5) \gr                  11) div B
  6) T
-------------------------------------------------------
 12) multiplot [  4 ]      m) set multiplot
-------------------------------------------------------
 d(ata) p(age) o(pts) l(imits) le(g)end h(elp)
 r(ender) v(ector) x(sec/rotate) s,S(ave) q(uit)
-------------------------------------------------------
Please enter your selection now (y axis or option):

The simplest plot is of two quantities which are not both coordinates. For example, to plot density vs smoothing length, type

Please enter your selection now (y axis or option): 5
(x axis) (default=1): 4
 Graphics device/type (? to see list, default /xwin): /xw

The default= refers to the default value assigned if you just press the return key. The last prompt asks for the device to which output should be directed.

Hint

A full list of available graphics devices is given by typing ? at the prompt.

Some of the most useful devices are given in Commonly used graphics devices available in giza. In the above we have selected the X-window driver which means that the output is sent to the screen (provided X-windows is running), as demonstrated in the screenshot shown in Fig. 1.

Screenshot of simple two column plot to an X-window¶

Commonly used graphics devices available in giza¶
`/xw`, `/xwin`	X-Window (interactive)	`/png`	Portable Network Graphics (bitmap)
`/eps`	Encapsulated postscript (one file per page)	`/svg`	Scalable Vector Graphics
`/pdf`	PDF	`/null`	null device (no output)
`/ps`	Postscript (all pages in one file)

Interactive mode¶

Many useful tasks can now be achieved by moving the mouse to the plot window and selecting areas or pressing keystrokes – this is Interactive mode. Most useful are:

press l with the mouse over the colour bar for a log axis
a to adapt the plot limits (with mouse on the colour bar, inside the plot, or positioned next to the x or y axes)
left click to select an area with the mouse and click to zoom
left click on the colour bar to change the rendering limits
space to skip to the next file (right click or b to go back)
- or + to zoom in or out
Enter for Hollywood mode
o to recentre the plot on the origin
r to refresh the plot (e.g. after changing the window size)
g to plot a line and find its gradient
m or M to change the colour map
f to flip the rendering to the next quantity
<, >, {, } and /, \ to rotate particles around z, y and x axes, respectively
G to move the legend
ctrl-t to annotate with text
backspace to delete annotation
s in the plot window to save changes between timesteps, otherwise the settings will revert when you move to the next timestep.
q in the plot window to quit the plotting window and return to the menu
q again from the splash main menu to quit splash altogether.
h in the plot window for the full list of commands

On particle plots you can additionally:

select an area and press 0-9 to colour particles (particle colours stick between plots, so you can use this to find particles with unusual parameters)
select an area and press 0 to hide selected particles
move the mouse over a particle and press c to see the size of the smoothing circle for that particle

These tasks can also be achieved non-interactively by a series of text-based Menus (see Changing plot settings). For example limits changing options are contained in the (l)imits menu, so to manually set plot limits we would type l from the main menu, then 2 for option 2 (set manual limits) and follow the prompts to set the limits for a particular data column. See also (i)nteractive mode.

Rendered plots¶

A more complicated plot is where both the $x-$ and $y-$ axes refer to coordinates. For example

Please enter your selection now (y axis or option):2
(x axis) (default=1): 1
(render) (0=none) ([0:11], default=0):5
(vector plot) (0=none, 8=B) ([0:8], default=0):0
Graphics device/type (? to see list, default /xwin): /xw

Notice that in this case that options appeared for rendered and vector plots. Our choice of 5 at the (render) prompt corresponds to column 5, which in this case is the density, producing the plot shown in the screenshot in Fig. 2.

Screenshot of 3D column density plot to an X-window¶

Important

Rendered plots only work if columns for density, particle mass and smoothing length are correctly identified in the data, and provided the number of coordinates is 2 or greater. Without these, rendering will just colour the points according to the selected column. See Data reads and command line options for internal details.

Cross section¶

To plot a cross section slice instead of a projection in 3D, type x at the main menu to open the (x) cross section/3D plotting options and choose option 1) switch between cross section and projection. Then re-plot the rendered plot again (exactly as in the previous example Rendered plots), setting the slice position at the prompt:

enter z position for cross-section slice: ([-8.328:8.327], default=0.000):

which produces the plot shown in the screenshot in Fig. 3

Screenshot of 3D Cross section slice plot to an X-window¶

Vector plots¶

A prompt to plot vector arrows on top of Rendered plots (or on top of particle plots) appears whenever vectors are present in the data (for details of how to specify this in your data read, see Data reads and command line options), taking the form:

(vector plot) (0=none, 8=B) ([0:8], default=0):0

where the number refers to the column of the first component of the vector quantity.

Vector plots in 3D show either the integral of each component along the line of sight or, for a Cross section, the vector arrows in a Cross section slice (depending on whether a projection or Cross section has been selected for 3D plots – see the rendering examples given previously). In 2D vector plots simply show the vector arrows mapped to a pixel array using the SPH kernel.

Settings related to vector plots can be changed via (v)ector plot options. The size of the arrows is set by the maximum plot limit over all of the vector components. Alternatively the arrow size can be changed interactively using v, V (decrease /increase the arrow size) and w (automatically adjust arrow size so longest arrow is one pixel width).

Contour plots¶

To plot contours of a quantity instead of Rendered plots, simply set the colour scheme used for rendering to 0 (contours only) via the “change colour scheme” option in the (r)endering options (type “r2” from the main menu as a shortcut to option 2 in the (r)endering options).

Contours of an additional quantity can also be plotted on top of Rendered plots. However the prompt for an additional contour plot does not appear by default – it can be turned on via the plot contours option in the (r)endering options (type r3 at the main menu as a shortcut). With this option set and a non-zero response to the render prompt, a prompt appears below the render prompt:

(render) (0=none) ([0:11], default=0):5
(contours) (0=none) ([0:11], default=0):6

Entering the column to use in the contour plot at this prompt (e.g. column 6 in the above example would correspond to the temperature) gives a rendered plot with overlaid contours.

Entering the same quantity used in the rendering at this prompt (e.g. column 5 in the above example) triggers a subsequent prompt for the contour limits which can then be set differently to those used in the render plot. In this way it is possible to make a plot where the density of one particle type is shown by the rendered plot and the density of another particle type (with different limits) is shown by contours. This can be achieved because once contour plotting is turned on, the contribution of a given particle type to either the contours or rendered plots can be turned on or off via the turn on/off particles by type option in the particle plot (o)ptions.

Moving forwards and backwards through data files¶

See Interactive mode. If you have put more than one file on the command line (or alternatively the file contains more than one dump), it is then possible to move forwards and backwards through the data:

press the space bar to move to the next file (with the cursor in the plot window - this is Interactive mode).
press b to load and plot the previous file
type 9 and press space to move forward by 9 files
type 10 and press b to move back by 10 files

Press h in Interactive mode for more.

Important

If you plot to a non-interactive device, splash simply cycles through all the files on the command line automatically.

Zooming in and out / changing plot limits¶

See Interactive mode. Having plotted to an interactive device (e.g. /xw), tasks such as zooming in and out, selecting, colouring and hiding particles, changing the limits of both the plot and the colour bar and many other things can be achieved using either the mouse (i.e., selecting an area on which to zoom in) or by a combination of the mouse and a keystroke.

Producing figures for LaTeX documents¶

Producing a pdf or postscript plot suitable for inclusion in a LaTeX file is simple. At the device prompt, type

Graphics device/type (? to see list, default /xw): myfile.eps

that is, instead of /xw (for an X-window), simply type /eps or .eps to use the encapsulated postscript driver. This produces a file which by default is called splash.eps, or if multiple files have been read, a sequence of files called splash_0000.eps, splash_0001.eps, etc. To specify both the device and filename, type the full filename (e.g. myfile.eps) as the device. Files produced in this way can be directly incorporated into LaTeX using standard packages.

Danger

Do not use the /png driver to produce files for LaTeX documents. Your axes will appear pixellated and blurred. Use a vector graphics device (eps or pdf) instead. These give clean, sharp and infinitely scalable text and lines.

Hint

Using eps format is recommended for LaTeX as it will always crop to the exact boundaries of the plot. The inbuilt pdf driver may require cropping of whitespace. Encapsulated postscript can be easily converted to pdf (for pdflatex) on the command line using:

epstopdf file.eps

Most pdflatex engines (including Overleaf) will handle/convert eps automatically.

Producing a sequence of plots for a movie¶

To make a movie of your simulation, first specify all of the files you want to use on the command line:

splash dump_*

and use an interactive device to adjust options until it looks right.

Hint

Movies look best with minimal annotation, e.g. using Hollywood mode or the backspace key in interactive mode to manually delete annotation

If in interactive mode type s to save the current settings, then plot the same thing again but to a non-interactive device. For example, to generate a sequence of png files

Graphics device/type (? to see list, default /xw): /png

This will generate a series of images named splash_0000.png, splash_0001.png, splash_0002.png corresponding to each new plotting page generated (or enter “myfile.png” at the device prompt to generate myfile_0000.png, myfile_0001.png, myfile_0002.png…).

Hint

Avoid prompts altogether using the Command line options. For example, to produce the above sequence of files from the command line, use

splash -r 5 -dev /png

Producing a movie from a sequence of images¶

Having obtained a sequence of images there are a variety of ways to make these into an animation using both free and commercial software. The simplest is to use ffmpeg:

ffmpeg -i splash_%04d.png -r 10 -vb 50M -bt 100M -pix_fmt yuv420p -vf setpts=4.*PTS movie.mp4

A simple script which executes the above command is included in the source file distribution

~/splash/scripts/movie.sh

See How do I make a movie from splash output? for more.

Hollywood mode¶

Press Enter or ctrl-m in the interactive plot window to start Hollywood mode, which changes to plot settings better suited to movies.

The following shows A circumbinary disc simulation viewed in default mode:

A circumbinary disc simulation viewed in default mode¶

and The same simulation viewed in Hollywood mode

The same simulation viewed in Hollywood mode¶

Remote visualisation¶

Visualisation of data in-situ on a cluster or supercomputer is simple. Just log in using ssh with X-windows forwarding, e.g.:

ssh -Y dprice@gadi.nci.org.au

Then just plot to an interactive device (/xw) as usual and everything in Interactive mode should just work.

splash has few dependencies and is simple to install in your home space if necessary. That said, it is always a good idea to get admins to install a shared package for all users.

Ten quick hints for producing good-looking plots¶

These can improve the look of a visualisation substantially compared to the default options:

Log the colour bar. To do this simply move the cursor over the colour bar and hit l (for log) in Interactive mode. Or non-interactively via the apply log or inverse transformations to columns option in the (l)imits menu.
Adjust the colour bar limits. Position the mouse over the colour bar and left-click in Interactive mode. To revert to the widest max/min possible for the data plotted, press a with the cursor positioned over the colour bar. Limits can also be set manually in the (l)imits menu.
Change the colour scheme. Press m or M in Interactive mode to cycle forwards or backwards through the available colour schemes.
Change the paper size. To produce high-resolution images/movies, use the change paper size option in the (p)age options to set the paper size in pixels.
Try normalised or exact interpolation. If your simulation does not involve free surfaces (or alternatively if the free surfaces are not visible in the figure), turning the normalise interpolations option on (in the (r)endering options) may improve the smoothness of the rendering. This is turned off by default because it leads to funny-looking edges. Exact rendering performs exact sub-pixel rendering so is more accurate but slower.
Remove annotation/axes. For movies, often axes are unnecessary and detract from the visual appeal. Use Hollywood mode or delete annotation by pressing backspace in Interactive mode. Alternatively each can be turned off manually – axes via the axes options option in the (p)age options; the colour bar by the colour bar options entry in the (r)endering options and the legends via options in the le(g)end and title options.
Change axes/page colours. The background colour (colour of the page) and foreground colour (used for axes etc) can be changed vie the set foreground/background colours option in the (p)age options.
Move the legend or turn it off. The time legend can be moved by positioning the mouse and pressing G in interactive mode. The legend can be turned off in the le(g)end and title options or by pressing backspace in interactive mode. Similarly the vector plot legend can be turned on/off in the (v)ector plot options and moved by positioning the cursor and pressing H.
Use physical units on the axes. These can be set via the (d)ata options. See Changing physical unit settings for more details.
Save settings to disk! Don’t waste your effort without being able to reproduce the plot you have been working on. Pressing s in interactive mode only saves the current settings for subsequent timesteps. Pressing s from the main menu saves these settings to disk. Pressing S from the main menu saves both the plot options and the plot limits, so that the current plot can be reproduced exactly when splash is next invoked. Adding an a, as in SA, Sa or sa to the save options gives a prompt for a different prefix to the filenames (e.g. splash.defaults becomes myplot.defaults), which splash can be invoked to use via the -p command line option (e.g. splash -p myplot file1 file2...).

Changing plot settings¶

Menus¶

The plot settings may be changed in a series of submenus:

Options can be saved using the (s)ave option from the menu:

(s)aving your settings¶

Pressing s saves current options to splash.defaults.
Delete splash.defaults to revert all settings
Pressing S saves both splash.defaults and splash.limits (and any other files).
Typing sa or Sa gives a “save-as” option, changing the prefix of saved files
read defaults files saved with a different prefix using the -p option

Example:

Please enter your selection now (y axis or option):SA
enter prefix for filenames: (default="blah"):
default options saved to file blah.defaults
saving plot limits to file blah.limits
saving units to blah.units

Read this back using:

splash -p blah

Files saved by splash can also be edited manually. For example, splash.limits is a simple two-column ascii file containing minimum and maximum plot limits for each column. To reset the plot limits simply delete the splash.limits file.

set (m)ultiplot¶

Plotting more than one column from the same file on the same page (multiplot)¶

Press ’m’ (set (m)ultiplot) from the main menu to set up a multiplot.

Once you have gone through the options to set up a multiplot, to actually plot what you have set simply type the number of the column corresponding to “multiplot” at the $y-$axis prompt.

Important

A “multiplot” - multiple columns plotted from the same file - is different to plotting “multiple plots per page” - divide the plotting page up into panels. The number of panels across and down on a page can be changed (see Dividing the plotting page into panels) irrespective of whether or not you are also plotting multiple columns from the same file.

Plotting each particle type in a different panel (multiplot)¶

To make a plot using different particle types in each panel (e.g. gas density in one panel, dust or dark matter density in another), use ’m’ set (m)ultiplot from the main menu. If multiple types are present in the data read, the option appears to specify the particular types you want to use for each plot.

For example, after pressing ‘m’ at the main menu we eventually arrive at the question:

use all active particle types? (default=yes): n

Answering no brings up a possible list of types:

use gas particles
use ghost particles
use sink particles
use star particles
use unknown/dead particles
Enter type or list of types to use ([1:5], default=1): 1,3

Thus entering e.g. 1,3 specifies that only gas and sink particles should be used for this plot.

Important

This is more specific than simply turning particle types on and off for all plots, which can be achieved via the turn on/off particles by type option in the particle plot (o)ptions (see Plotting non-gas particles (e.g. ghosts, boundary, sink particles)).

(d)ata options¶

The following can all be achieved from the d)ata options menu:

Re-reading the initial data / changing the dump file¶

Just exit splash and restart with the new dump file name on the command line (remember to save by pressing ’S’ from the main menu before exiting to save both the current settings and the plot limits – then you can continue plotting with the current settings using a new dump file).

If you have placed more than one file on the command line, then pressing space in Interactive mode will read (and plot) the next file (press ’h’ in Interactive mode for a full list of commands - you can move forwards and backwards using arbitrary jumps). For non-interactive devices or where Interactive mode is turned off dump files are cycled through automatically, plotting the same plot for each file/timestep.

Using only a subset of data files / plotting every $n-$th dump file¶

To plot a subset of the data files in *any* order, see sec:selectedstepsonly.

Of course, another way to achieve the same thing is to explicitly order the files on the command line. A method I often use is to write all filenames to a file, e.g.

> ls DUMP* > splash.filenames

then edit the file to list only the files I want to use, then invoke splash with no files on the command line:

> splash

which will use the list of files specified in the splash.filenames file.

Plotting more than one file without re-reading the data from disk¶

For small data sets (or a small number of dump files) it is often useful to read all of the data into memory so that you can move rapidly forwards and backwards between dumps (e.g. in Interactive mode, or where both dumps are plotted on the same page) without unnecessary re-reading of data from disk. This is achieved with the command line flag:

splash --buffer file_0*

(provided you have the memory of course!!). Non-buffered data means that only one file at a time is read.

Calculating additional quantities not dumped¶

Turn calculate extra quantities on in the (d)ata options. New columns of data can be created as completely arbitrary functions of the data read from the SPH particles. Option d1 in the data menu leads, for a typical data read, to a prompt similar to the following:

 Specify a function to calculate from the data
 Valid variables are the column labels, 't', 'gamma', 'x0', 'y0' and 'z0' (origin setting)
 Spaces, escape sequences (\d) and units labels are removed from variable names
 Note that previously calculated quantities can be used in subsequent calculations

 Examples based on current data:
           r = sqrt((x-x0)**2 + (y-y0)**2 + (z-z0)**2)
           pressure = (gamma-1)*density*u
           |v| = sqrt(vx**2 + vy**2 + vz**2)

Enter function string to calculate (blank for none) (default=""):

Thus, one can for example calculate the pressure from the density and thermal energy according by copying the second example given.

Hint

Function calculation is completely general and can use any of the columns read from the file, the time for each step (‘t’), the adiabatic index $\gamma$ (‘gamma’) and the current origin setting (x0, y0 and z0). Previously calculated quantities can also be used - e.g. in the above example we could further compute, say, an entropy variable using s=pressure/density^gamma after the pressure has been specified. The resultant quantities appear in the main splash menu as standard columns just as if they had been read from the original data file.

The origin for the calculation of radius can be changed via the rotation on/off/settings option in the (x) cross section/3D plotting options. If particle tracking limits are set (see Making plot limits relative to a particular particle) the radius is calculated relative to the particle being tracked.

If you simply want to multiply a column by a fixed number (e.g. say you have sound speed squared and you want to plot temperature) - this can also be achieved by defining a unit for the column (i.e., a factor by which to multiply the column by) – see Plotting data in physical units for details. The corresponding label can be changed by creating a splash.columns file (or for the ascii read just a file called ‘columns’) containing labels which are used to override the default ones from the data read (one per line) – see Changing the default column labels for more details.

See also Plotting in different coordinate systems (e.g. cylindrical coordinates) for how to transform vectors (and positions) into different coordinate systems.

Plotting data in physical units¶

Data can be plotted in physical units by turning on the use physical units option in the (d)ata options. The settings for transforming the data into physical units may be changed via the change physical unit settings option in the (d)ata options. (see Changing physical unit settings)

For some data reads (phantom, sphNG, magma) the scalings required to transform the data into physical units are read from the dump file. These are used as the default values but are overridden as soon as changes are made by the user (that is, by the presence of a ‘splash.units’ file) (see Changing physical unit settings).

Important

Physical units are now ON by default in v3.x of SPLASH if they are set by the data read. You can use this option to revert to code units

Rescaling data columns¶

See Plotting data in physical units.

Changing the default column labels¶

The labelling of columns is usually automatic in the data format read (for ascii files labels will be read from the file header). Aside from changing the labels in the read_data file specific to the format you are reading, it is also possible to override the labelling of columns at runtime by creating a file called splash.columns (or with a different prefix if the -p command line option is used), with one label per line corresponding to each column read from the dump file, e.g.

column 1
column 2
column 3
my quantity
another quantity

Warning

Labels in the splash.columns file will not override the labels of coordinate axes or labels for vector quantities (as these require the ability to be changed by plotting in different coordinate systems – see Plotting in different coordinate systems (e.g. cylindrical coordinates)).

Plotting column density in g/cm$^{2}$ without having x,y,z in cm¶

See Changing physical unit settings. In addition to units for each column (and a unit for time – see Changing the time units) a unit can be set for the length scale added in 3D column integrated plots. The prompt for this appears after the units of either $x$, $y$, $z$ or $h$ has been changed via the change physical unit settings option in the (d)ata options. The length unit for integration is saved in the first row of the splash.units file, after the units for time.

See Changing the label used for 3D projection plots for details on changing the default labelling scheme for 3D column integrated (projection) plots.

Changing physical unit settings¶

The settings for transforming the data into physical units may be changed via the change physical unit settings option in the (d)ata options. To apply the physical units to the data select the use physical units option in the (d)ata options.

The transformation used is $new= old*units$ where old is the data as read from the dump file and new is the value actually plotted. The data menu option also prompts for a units label which is appended to the usual label. Brackets and spaces should be explicitly included in the label as required.

Once units have been changed, the user is prompted to save the unit settings to a file called splash.units. Another way of changing units is simply to edit this file yourself in any text editor (the format is fairly self-explanatory). To revert to the default unit settings simply delete this file. To revert to code units turn use physical units off in the (d)ata options.

Hint

A further example of where this option can be useful is where the $y-$axis looks crowded because the numeric axis labels read something like $1\times 10^{-4}$. The units option can be used to rescale the data so that the numeric label reads $1$ (by setting $units=10^{4}$) whilst the label string is amended to read $y [\times 10^{-4}]$ by setting the units label to $[ \times 10^{-4}]$.

Changing the axis label to something like $x$ $[ \times 10^{4} ]$¶

See Changing physical unit settings.

Changing the time units¶

Units for the time used in the legend can be changed using the change physical unit settings in the (d)ata options. Changing the units of column zero corresponds to the time (appears as the first row in the ‘splash.units’ file).

(i)nteractive mode¶

The menu option i) turns on/off Interactive mode. With this option turned on (the default) and an appropriate device selected (i.e., the X-window, not /png or /ps), after each plot the program waits for specific commands from the user. With the cursor positioned anywhere in the plot window (but not outside it!), many different commands can be invoked. Some functions you may find useful are: Move through timesteps by pressing space (b to go back); zoom/select particles by selecting an area with the mouse; rotate the particles by using the $<$, $>$,[, ] and $\backslash$, / keys; log the axes by holding the cursor over the appropriate axis and pressing the l key. Press q in the plot window to quit Interactive mode.

A full list of these commands is obtained by holding the cursor in the plot window and pressing the ‘h’ key (h for help).

Hint

Changes made in Interactive mode will only be saved by pressing the ‘s’ (for save) key. Otherwise pressing space (to advance to the next timestep) erases the changes made during Interactive mode. A more limited Interactive mode applies when there is more than one plot per page.

Many more commands could be added to the Interactive mode, limited only by your imagination. Please send me your suggestions!

Adapting the plot limits¶

Press a in Interactive mode to adapt the plot limits to the current minimum and maximum of the quantity being plotted. With the mouse over the colour bar, this applies to the colour bar limits. Also works even when the page is subdivided into panels. To adapt the size of the arrows on Vector plots, press w. To use “adaptive plot limits” (where the limits change at every timestep), see Using plot limits which adapt automatically for each new plot.

Making the axes logarithmic¶

Press l in Interactive mode with the mouse over either the x or y axis or the colour bar to use a logarithmic axis. Pressing l again changes back to linear axes. To use logarithmic labels as well as logarithmic axes, see Using logarithmic axes labels.

Cycling through data columns interactively¶

Use f in Interactive mode on Rendered plots to interactively ‘flip’ forwards to the next quantity in the data columns (e.g. thermal energy instead of density). Use ’F’ to flip backwards.

Colouring a subset of the particles and retaining this colour through other timesteps¶

Example of particles coloured interactively using the mouse (left) and selection using a parameter range (right), which is the same as the plot on the left but showing only particles in a particular density range (after an intermediate plot of density vs x on which I selected a subset of particles and hit p)¶

In Interactive mode, select a subset of the particles using the mouse (that is left click and resize the box until it contains the region you require), then press either 1-9 to colour the selected particles with colours corresponding to plotting library colour indices 1-9, press p to plot only those particles selected (hiding all other particles), or h to hide the selected particles. An example is shown in the left panel of Fig. 6. Particles retain these colours between timesteps and even between plots. This feature can therefore be used to find particles within a certain parameter range (e.g. by plotting density with x, selecting/colouring particles in a given density range, then plotting x vs y in which the particles will appear as previously selected/coloured). An example of this feature is shown in the right panel of Fig. 6 where I have plotted an intermediate plot of density vs x on which I selected a subset of particles and hit p (to plot only that subset), then re-plotted x vs y with the new particle selections.

To “un-hide” or “de-colour” particles, simply select the entire plotting area and press 1 to restore all particles to the foreground colour index.

Particles hidden in this manner are also no longer used in the rendering calculation. Thus it is possible to render using only a subset of the particles (e.g. using only half of a box, or only high density particles). An example is shown in Fig. 7.

To colour the particles according to the value of a particular quantity, see Using coloured particles instead of rendering to pixels.

Important

Selection in this way is based on the particle identity, meaning that the parameter range itself is not preserved for subsequent timesteps, but rather the subset of particles selected from the initial timestep. This can be useful for working out which particles formed a particular object in a simulation by selecting only particles in that object at the end time, and moving backwards through timesteps retaining that selection.

Working out which particles formed a particular object in a simulation¶

This can be achieved by selecting and colouring particles at a particular timestep and plotting the same selection at an earlier time. See Colouring a subset of the particles and retaining this colour through other timesteps for details.

Plotting only a subset of the particles¶

To turn plotting of certain particle types on and off, see Plotting non-gas particles (e.g. ghosts, boundary, sink particles). To select a subset of the particles based on restrictions of a particular parameter or by spatial region see Colouring a subset of the particles and retaining this colour through other timesteps.

Rendering using only a subset of the particles¶

Particles can be selected and ‘hidden’ interactively (see Colouring a subset of the particles and retaining this colour through other timesteps) – for Rendered plots ‘hidden’ particles are also not used in the interpolation calculation from the particles to the pixel array. An example is shown in Fig. 7, where I have taken one of the rendered examples in Basic splash usage, selected half of the domain with the mouse and pressed ’p’ to plot only the selected particles. The result is the plot shown.

Example of Rendered plots using only a subset of the particles. Here I have selected only particles on the right hand side of the plot using the mouse and hit ’p’ to plot only those particles.¶

Important

Selection of data subsets is by default based on particle identity – the same particles will be used for the plot in subsequent dumps, allowing one to easily track the Lagrangian evolution of a patch of gas. See Using a subset of data restricted by parameter range to select by fixed parameter ranges (e.g. to only show particles in fixed density range).

Hint

A range restriction can be set in Interactive mode by selecting the restricted box using the mouse and pressing x, y or r to restrict the particles used to the x, y (or r for both x and y) range of the selected box respectively. Pressing S at the main menu will save such range restrictions to the splash.limits file.

Tracking a set of particles through multiple timesteps¶

See Rendering using only a subset of the particles.

Taking an oblique cross section interactively¶

It is possible to take an oblique Cross section through 3D data using a combination of rotation and Cross section slice plotting. To set the position interactively, press x in Interactive mode to draw the position of the Cross section line (e.g. on an x-y plot this then produces a z-x plot with the appropriate amount of rotation to give the cross section slice in the position selected).

Hint

Interactively selecting a Cross section will work even if the current plot is a 3D column integrated projection. In this case the setting projection or cross section changes to cross section in order to plot the slice.

(p)age options¶

Options related to the page setup are changed in the p)age submenu.

Overlaying timesteps/multiple dump files on top of each other¶

It is possible to over-plot data from one file on top of data from another using the plot n steps on top of each other option from the (p)age options. Setting $n$ to a number greater than one means that the page is not changed until $n$ steps have been plotted. Following the prompts, it is possible to change the colour of all particles between steps and the graph markers used and plot an associated legend (see below).

Hint

This option can also be used in combination with a multiplot (see set (m)ultiplot) – for example plotting the density vs x and pressure vs x in separate panels, then with $n > 1$ all timesteps will be plotted in each panel.

When more than one timestep is plotted per page with different markers/colours, an additional legend can be plotted (turn this on in the le(g)end and title options, or when prompted while setting the plot n steps on top of each other option). The text for this legend is just the filename by default, if one timestep per file, or just something dull like ’step 1’, if more than one timestep per file.

Hint

To change the legend text, create a file called splash.legend in the working directory, with one label per line. The position of the legend can be changed either manually via the legend and title options in the (p)age options, or by positioning the mouse in Interactive mode and pressing G (similar keys apply for moving plot titles and the legend for Vector plots – press h in Interactive mode for a full list).

Plotting results from multiple files in the same panel¶

See Overlaying timesteps/multiple dump files on top of each other.

Plotting more than one dump file on the same page¶

This is slightly different to “plotting more than one dump file on the same panel”

Changing axes settings¶

Axes settings can be changed in the (p)age options, by choosing axes options. The options are as follows:

-4 : draw box and major tick marks only;
-3 : draw box and tick marks (major and minor) only;
-2 : draw no box, axes or labels;
-1 : draw box only;
 0 : draw box and label it with coordinates;
 1 : same as AXIS=0, but also draw the coordinate axes (X=0, Y=0);
 2 : same as AXIS=1, but also draw grid lines at major increments of the coordinates;
 3 : draw box, ticks and numbers but no axes labels;
 4 : same as AXIS=0, but with a second y-axis scaled and labelled differently
10 : draw box and label X-axis logarithmically;
20 : draw box and label Y-axis logarithmically;
30 : draw box and label both axes logarithmically.

Turning axes off¶

Plot axes can be turned off by choosing axes options in the (p)age options or by deleting them using the backspace key in Interactive mode. See Changing axes settings for more details.

Turning axes labels off¶

Axes labels and numbering can be turned off via the axes options option in the (p)age options or by deleting them using the backspace key in Interactive mode. See Changing axes settings for more details.

Using logarithmic axes labels¶

Logarithmic axes (that is where the quantity plotted is logged) can be set via the apply log or inverse transformations option in the (l)imits menu or simply by pressing l with the cursor over the desired axis (or the colour bar) in Interactive mode. By default the axes labels reads $log(x)$ and the number next to the axis is $-4$ when $x$ is 10$^{-4}$. Logarithmic axes labels (i.e., where the label reads $x$ and the number next to the axis is $10^{-4}$ with a logarithmic scale) can be specified by choosing the axes options option in the (p)age options and setting the axes option to 10, 20 or 30 as necessary (see Changing axes settings for more details).

Plotting a second, rescaled y-axis on the right hand side of a plot¶

A second y axis can be added by selecting the axis=4 option in the axes option in the (p)age options (see Changing axes settings). This will prompt for the scaling and alternative label:

enter axis option ([-4:30], default=0): 4
enter scale factor for alternative y axis ([0.000:], default=1.000): 10.0
enter label for alternative y axis (default=""): y [other units]

Changing the size of the plotting surface¶

The physical size of the viewing surface used for plotting can be changed via the change paper size option in the (p)age options. This affects the size of the X-window (if plotted to the screen) and the size of .png or images generated (if plotted to these devices). Several preset options are provided or the paper size in x and y can be explicitly specified in inches or pixels.

Dividing the plotting page into panels¶

The plotting page can be divided into panels using the subdivide page into panels option in the (p)age options. For multiple plots per page (i.e., nacross $\times$ ndown $> 1$) a more limited Interactive mode applies (basically because the data used for the plots is no longer stored in memory if there is more than one plot on the same page meaning that functionality such as selecting particles must be turned off).

Tiling plots with the same $x-$ and $y-$ axes¶

Plots with the same $x-$ and $y-$ axes are tiled if the tiling option from the (p)age options. Tiling means that only one axis is shown where multiple plots share the same x or y axis and that the plots are placed as close to each other as possible. For Rendered plots a shared colour bar is plotted which spans the full length of the page.

Using non-proportional scales for spatial dimensions¶

By default if the x and y axes are both spatial coordinates, the axes are scaled proportionately. This can be changed via the spatial dimensions have same scale option in (p)age options.

Using non-square axes on coordinate plots¶

See Using non-proportional scales for spatial dimensions.

Changing the character height for axes, labels and legends¶

The character height used for axes, labels and legends can be changed via the (p)age options. The character height is relative to the paper size where 1.0 = 1/40th of the page height. The page height can be changed in Changing the size of the plotting surface.

Using a thicker line width on plots¶

The line width used for axes and text can be changed via the (p)age options.

Changing the foreground and background colours¶

The background and foreground colour of a plot can be changed via the set foreground/background colours option in the (p)age options. Note that the background colour setting has no effect on postscript devices (see Producing figures for LaTeX documents for more details).

Plotting axes, legends and titles in white even when the labels are plotted in black¶

By default, axes, legends and titles are plotted in the foreground colour (e.g. black). However if the plot itself is also largely black (e.g. Rendered plots when lots of particles are plotted) it can be useful to overplot those parts of the axes and labelling which lie on top of the plotting surface in the background colour (e.g. white). A prompt for this is given when setting the set foreground/background colours option in the (p)age options.

The prompt appears as follows:

---------------- page setup options -------------------
...
 9) set foreground/background colours
enter option ([0:8], default=0):9
 Enter background colour (by name, e.g. "black") (default=""):white
 Enter foreground colour (by name, e.g. "white") (default=""):black

 Overlaid (that is, drawn inside the plot borders) axis
 ticks, legend text and titles are by default plotted in
 the foreground colour [i.e., black].

Do you want to plot these in background colour [i.e., white] instead ? (default=no):y

In the above I have selected a background colour of white, a foreground colour of black. Answering yes to the last question means that those parts of the axes which lie on top of the viewing surface (and any labels) will be plotted in white (the background colour) instead of the foreground colour (black).

le(g)end and title options¶

Adding titles to plots / repositioning titles¶

Plots may be titled individually by creating a file called splash.titles in the current directory, with the title on each line corresponding to the position of the plot on the page. Thus the title is the same between timesteps unless the steps are plotted together on the same physical page. Leave blank lines for plots without titles. For example, creating a file called splash.titles in the current directory, containing the text:

plot one
plot two
plot three

and positioning the title using the default options, will produce a plot with one of these titles on each panel.

Turning off/moving the time legend¶

The position of the time legend can be set interactively by positioning the mouse in the plot window and pressing ’G’. To set the position non-interactively and/or change additional settings such as the justification, use the time legend on/off/settings option in the le(g)end and title options.

Changing the text in the time legend¶

The text which appears the time legend (by default this is t=) can be changed via the time legend on/off/settings option in the le(g)end and title options.

To rescale the value of the time displayed in the time legend (default value is as read from the dump file), see Changing the time units.

Making the legend read ‘z=’ instead of ‘t=’¶

See Changing the text in the time legend. An option to change the legend text is provided in the time legend on/off/settings option in le(g)end and title options. The numeric value of the time legend is as read into the time array in the read_data routine. This value can be rescaled by setting a unit for time (see Changing the time units).

Plotting the time legend on the first row/column of panels / nth panel only¶

An option to plot the time legend on the first row or column of panels or on a single panel only appears in the le(g)end and title options.

Plotting a length scale on coordinate plots¶

An option to plot a length scale (i.e., |---| with a label below it indicating the length) on coordinate plots (i.e., plots where both $x-$ and $y-$axes refer to particle coordinates) is provided in the le(g)end and title options.

Annotating a plot with squares, rectangles, arrows, circles and text¶

Use the annotate plot option in le(g)end and title options to annotate plots with a range of geometric objects (squares, rectangles, arrows, circles and text) with full control over attributes such as line width, line style, colour, angle and fill style.

Text annotation can also be added/deleted in Interactive mode using ctrl-t (to add) and the backspace key (to delete). Text can also be added to plots by adding titles (Adding titles to plots / repositioning titles) which can be different in different panels. Text labels added using shape annotation differ from titles by the fact that they must appear the same in each panel and are positioned according to the world co-ordinates of the plot (rather than relative to the viewport). Shape text can also be displayed at arbitrary angles.

An option to plot length scales (|---|) on coordinate plots is implemented separately via the plot scale on coordinate plots option in le(g)end and title options.

Adding your name to a plot/movie¶

Arbitrary text annotation can be added/removed in Interactive mode using ctrl-t (to add) and the backspace key (to delete) or via the annotate plot option in the le(g)end and title options.

particle plot (o)ptions¶

The following are tasks which can be achieved via options in the particle plot (o)ptions.

Plotting non-gas particles (e.g. ghosts, boundary, sink particles)¶

Particles of different types can be turned on or off (i.e., plotted or not) using the turn on/off particles by type option in particle plot (o)ptions. This option also prompts to allow particles of non-SPH types to be plotted on top of Rendered plots (useful for sink or star particles - this option does not apply to SPH particle types). Turning SPH particle types on or off also determines whether or not they will be used in the rendering calculation (i.e., the interpolation to pixels). This particularly applies to ghost particles, where ghost particles will only be used in the rendering if they are turned on via this menu option.

(The fact that particles of a given type are SPH particles or not is specified by the UseTypeInRendering flags in the set_labels part of the read_data file).

Plotting non-gas particles on top of rendered plots¶

An option to plot non-SPH particles on top of Rendered plots (e.g. sink particles) can be set when turning particle types on/off via the turn on/off particles by type option in particle plot (o)ptions (see Plotting non-gas particles (e.g. ghosts, boundary, sink particles)).

Using ghost particles in the rendering¶

See Plotting non-gas particles (e.g. ghosts, boundary, sink particles).

Turn off plotting of gas particles¶

Particles can be turned on or off by type via the turn on/off particles by type option in particle plot (o)ptions. See Plotting non-gas particles (e.g. ghosts, boundary, sink particles).

Plotting dark matter particles¶

To plot dark matter particles (e.g. for the gadget read) the particle type corresponding to dark matter particles must be turned on via the turn on/off particles by type option in particle plot (o)ptions. Turning this option on means that dark matter particles will appear on particle plots.

To make Rendered plots of dark matter (e.g. showing column density), it is necessary to define smoothing lengths and a fake “density” for the dark matter particles. If your data read already supplies individual smoothing lengths for dark matter particles, the only thing to do is define a fake density field with a constant value (e.g. $\rho = 1$ for all dark matter particles). The actual density value does not matter, so long as it is non-zero, as the rendering for density does not use it unless the normalise interpolations option in the (r)endering options is set (which it is not by default). This is because SPLASH constructs the weight:

\[w_{part} = \frac{m_{part}}{\rho_{part} h_{part}^{\nu}},\]

(see Price 2007) and then interpolates for any quantity A using

\[A_{pixels} = \sum_{part} w_{part} A_{part} W_{kernel},\]

so if $A = \rho$ then the actual rho value cancels.

For the GADGET data read you can define the smoothing length for dark matter particles by setting the environment variable GSPLASH_DARKMATTER_HSOFT (see sec:gsplash for details), which also triggers the creation of a fake density column as required. With this variable set dark matter particles are treated identically to SPH particles and can be rendered as usual (although the only meaningful quantity to render is the density). A much better way is to define smoothing lengths individually for dark matter particles, for example based on a local number density estimate from the relation

\[h \propto n^{-1/3}, \hspace{0.5cm} \textrm{where} \hspace{0.5cm} n_{i} = \sum_{j} W_{ij}.\]

Actually, none of this should be necessary, as the gravity for dark matter should be softened with smoothing lengths defined like this in the first place. The historical practice of fixed softening lengths has arisen only because of confusion about what softening really means (and worries about energy conservation with adaptive softening lengths). What you are trying to do is solve Poisson’s equation for the dark matter density field, defined with a kernel density estimate and using fixed softening lengths is not a way to get a good density… but don’t get me started, read [PM07] instead.

Danger

For simulations using both SPH and dark matter particles, dark matter particles will contribute (incorrectly) to the SPH rendering when the environment variable is set and the plotting of dark matter particles is turned on. Thus to plot just gas column density in this case, dark matter particles must be turned off [via the particle plot (o)ptions], and similarly to plot just dark matter density if both SPH and dark matter particles are present, SPH particles must be turned off.

Plotting a column density plot of dark matter/N-body particles¶

See Plotting dark matter particles.

Plotting sink particles¶

Sink particles will be plotted on particle plots once turned on via the turn on/off particles by type option in particle plot (o)ptions. Setting this option also gives a prompt for whether or not to plot sink particles on top of Rendered plots (to which the answer should be yes). See Plotting non-gas particles (e.g. ghosts, boundary, sink particles) for more details.

To plot sink particles as a circle scaled to the sink radius, select the appropriate marker type (32-35) in the change graph markers for each type option in particle plot (o)ptions. This allows plotting of particles of a given type with circles, filled or open, proportional to their smoothing lengths. Thus, the smoothing length for sink particles needs to be set to their accretion radius (or at least proportional to it).

Hint

A good option for sink particles is to print “outlined” filled circles (marker 34) — these show up on both black or white backgrounds.

Plotting sink particles with size proportional to the sink radius¶

See Plotting sink particles.

Plotting a point mass particle with physical size¶

See Plotting sink particles.

Changing graph markers for each particle type¶

The graph markers used to plot each particle type can be changed via the change graph markers for each type option in the particle plot (o)ptions. The full list of available markers is given in the documentation for giza (also similar to the markers used in pgplot).

SPLASH also allows the particles to be marked by a circle proportional to the smoothing length for that particle, implemented as marker types 32-35 under the change graph markers for each type option in the particle plot (o)ptions.

Plotting each particle type in a different colour¶

Each particle type can be plotted in a different colour via the set colour for each particle type option in the particle plot (o)ptions.

Changing the order in which different particle types are plotted¶

The order in which particle types are plotted can be changed via the change plotting order of types option in particle plot (o)ptions. For example, it is possible to make dark matter particles be plotted on top of gas particles rather than the default which is vice-versa. This is only implemented for particle types which are stored contiguously (one after the other) in the data read, rather than mixed in with each other.

Plotting using lines instead of dots (e.g. for energy vs time plots)¶

An option to plot a line joining all of the points on a plot can be set via the plot line joining particles option in particle plot (o)ptions. When set, this option plots a line connecting the (gas only) particles in the order that they appear in the data array. Useful mainly in one dimension or when plotting ascii data, although can give an indication of the relative closeness of the particles in memory and in physical space in higher dimensions. The line colours and styles can be changed.

To plot the line only with no particles, turn off gas particles using the turn on/off particles by type option from particle plot (o)ptions.

Plotting multiple lines with different colours/line styles and a legend¶

When multiple timesteps are plotted on the same physical page, the line style can be changed instead of the colour (this occurs when the change colour option is chosen for multiple steps per page – see the change plots per page option in the (p)age options.

Joining the dots¶

See Plotting using lines instead of dots (e.g. for energy vs time plots).

Plotting the size of the smoothing circle around selected particles¶

On coordinate plots this option plots a circle of radius $2h$ around selected particles. This is primarily useful in debugging neighbour finding routines. Where only one of the axes is a coordinate this function plots an error bar of length $2h$ in either direction is plotted in the direction of the coordinate axis. See also Locating a particular particle in the data set for more details.

Locating a particular particle in the data set¶

The best way to locate a particular particle in the data set is to use the plot smoothing circles option in particle plot (o)ptions, e.g:

Please enter your selection now (y axis or option):o5
------------- particle plot options -------------------
 Note that circles of interaction can also be set interactively
Enter number of circles to draw ([0:100], default=0):1
Enter particle number to plot circle around ([1:959], default=1): 868

then upon plotting a coordinate plot (e.g. x vs y), particle 868 will be plotted with a circle of size $2h$ which makes it easy to distinguish from the other particles. See also Plotting the size of the smoothing circle around selected particles.

Plotting in different coordinate systems (e.g. cylindrical coordinates)¶

The coordinates of position and of all vector components can be transformed into non-cartesian coordinate systems using the change coordinate system option in particle plot (o)ptions. For example, a dump file with columns as follows:

-------------------------------------------------------
  1) x                     6) log density
  2) y                     7) v\dx
  3) z                     8) v\dy
  4) particle mass         9) v\dz
  5) h
-------------------------------------------------------
 10) multiplot [  4 ]      m) set multiplot
-------------------------------------------------------
Please enter your selection now (y axis or option):

choosing particle plot (o)ptions, option 7) and choosing cylindrical coordinates then produces;

 You may choose from a delectable sample of plots
-------------------------------------------------------
  1) r                     6) log density
  2) phi                   7) v\dr
  3) z                     8) v\dphi
  4) particle mass         9) v\dz
  5) h
-------------------------------------------------------
...

transforming both coordinates and vectors into the chosen coordinate system.

Hint

Rendered plots are also possible in coordinate systems other than those native to the file.

Hint

For 3D SPH simulations, extra columns will appear in the menu in cylindrical or spherical coordinates allowing plots of azimuthally-averaged surface density and Toomre Q parameter. For more details see Plotting azimuthally-averaged disc surface density and Toomre Q parameter.

Details of the coordinate transformations are given in Coordinate transforms.

If you have a coordinate system you would like implemented, please email me the details!

Plotting the exact solution to common test problems¶

Turn on plot exact solution in particle plot (o)ptions

The following exact solutions are provided

Any arbitrary function y = f(x,t) (can be plotted on any or all of the plots). The functions to be plotted can also be specified by creating a splash.func file with one function per line.
Hydrodynamic shock tubes (Riemann problem) – a full solution is provided for all types of waves propagating in either direction.
Spherically-symmetric 3D sedov blast wave problem.
Polytropes (with arbitrary $\gamma$)
One and two dimensional toy stars. This is a particularly simple test problem for SPH codes described in [MP04].
Linear wave. This simply plots a sine wave of a specified amplitude, period and wavelength on the plot specified.
MHD shock tubes (tabulated). These are tabulated solutions for 7 specific MHD shock tube problems.
h vs $\rho$. This is the exact solution relating smoothing length and density in the form $h \propto (m/\rho)^{1/\nu}$ where $\nu$ is the number of spatial dimensions.
radial density profiles. For various models commonly used in $N-$body simulations.
Exact solution from a file. This option reads in an exact solution from the filename input by the user, assuming the file contains two columns containing the $x-$ and $y-$ coordinates of an exact solution to be plotted as a line on the plot specified.

Details of the calculation of the exact solutions are given in Exact solutions. An example plot using the Sedov blast wave exact solution is shown in Fig. 8.

Example of a plot utilising the Sedov blast wave exact solution. Taken from [RP07].¶

Plotting an exact solution from a file¶

See Plotting the exact solution to common test problems. One of the options for exact solution plotting is to read the exact solution from either one or a sequence of ascii files, such that the results are plotted alongside the particle data. The filename(s) can be specified by the user and will be saved to the ‘splash.defaults’ file so that the solution(s) will be read and plotted on subsequent invocations of splash .

Changing the exact solution line style & colour¶

The line style and colour of the exact solution line can be changed via the exact solution plot options option in particle plot (o)ptions. This option can also be used to turn on/off calculation of various error norms together with an inset plot of the residual error on the particles. See Exact solutions for details of the error norms calculated.

Setting the number of points used in an exact solution calculation¶

The number of points used in an exact solution calculation can be changed via the exact solution plot options option in particle plot (o)ptions.

Plotting an inset plot of residual errors from an exact solution¶

An inset plot of residual errors between the plotted points and an exact solution calculation can be turned on via the exact solution plot options option in particle plot (o)ptions.

(l)imits menu¶

Using plot limits which adapt automatically for each new plot¶

Adaptive plot limits can be set using option 1 of the (l)imits menu. Different settings can be applied to coordinate axes and non-coordinate axes. Changing plot limits interactively and pressing s in Interactive mode will change this option back to using fixed limits.

Using adaptive plot limits for the colour bar but not for the coordinates¶

Adaptive plot limits can be set individually for coordinate axes and non-coordinate axes (e.g. the colour bar) via the use adaptive/fixed limits option in the (l)imits menu. See Using plot limits which adapt automatically for each new plot.

Setting plot limits manually¶

Plot limits can be set manually using option 2) of the (l)imits menu (or simply l2 from the main menu). Alternatively you can edit the splash.limits file created by (s)aving your settings prior to invoking splash (this file simply contains the minimum and maximum limits for each column on consecutive lines).

Making plot limits relative to a particular particle¶

Particle tracking limits (i.e., where a chosen particle is always at the centre of the plot and limits are set relative to that position) can be set via the make xy limits relative to particle option in the (l)imits menu. Alternatively particle tracking limits can be set interactively by pressing t in Interactive mode with the cursor over the particle you wish to track.

Warning

This option only works if particle identities are preserved between timesteps in the data files

Important

With particle tracking limits set, the radius calculated via the calculate extra quantities option in the (d)ata options is calculated relative to the tracked particle.

Centreing on a sink particle can also be achieved using the SPLASH_CENTRE_ON_SINK environment variable (see Environment variables).

Plotting in a comoving reference frame¶

A co-moving reference frame can be set using the make xy limits relative to particle option in the (l)imits menu. Coordinate limits are then centred on the selected particle for all timesteps, with offsets as input by the user. This effectively gives the ‘Lagrangian’ perspective. See Making plot limits relative to a particular particle for more details. Centreing on a sink particle can also be achieved using the SPLASH_CENTRE_ON_SINK environment variable.

Setting the origin to correspond to a particular particle¶

See Making plot limits relative to a particular particle.

Tracking a particle¶

See Making plot limits relative to a particular particle.

Setting the origin to the position of the $n$th sink particle¶

This can be achieved using the make xy limits relative to particle option in the (l)imits menu. For example, to track the first sink particle we would proceed as follows:

Please enter your selection now (y axis or option):l3
------------------ limits options ---------------------
To track particle 4923, enter 4923
To track the 43rd particle of type 3, enter 3:43

Enter particle to track: (default="0"): 3:1

where 3:1 indicates the first particle of type 3. The origin is set to the position of this particle and limits are relative to its position. See Making plot limits relative to a particular particle for more details.

Plotting radial plots around sink particles¶

First, set the origin to the location of the sink, as described above. Then simply change to spherical coordinates using the change coordinate systems option in particle plot (o)ptions. Alternatively, compute the radius using the calculate extra quantities option in the (d)ata options.

Automatically adapting plot limits to match aspect ratio of output device¶

An option to automatically adjust the plot limits to match the aspect ratio of the output device is given in the (l)imits menu, and is also prompted for whenever the paper size is changed (via the change paper size option in the (p)age options, see Changing the size of the plotting surface).

Plotting with log axes.¶

Log axes can be set either interactively (by pressing l with the cursor over the desired axis) or manually via the apply log or inverse transformations to columns option in the (l)imits menu. To use logarithmic axes labels as well, see Using logarithmic axes labels.

Plotting the square root, inverse or square of a quantity¶

Columns can be logged, inverted, sqrt-ed, squared or any combination of the above via the apply log or inverse transformations to columns option in the (l)imits menu. If you have any additional transformations you would find useful please let me know, as it is straightforward to add more.

Resetting limits for all columns¶

Limits for all columns can be reset to their minimum and maximum values from the current dump file via the reset limits for all columns option in the (l)imits menu. See Interactive mode for details of resetting plot limits for a particular plot interactively.

Restoring all plot limits to their minimum and maximum values in the current dump file¶

See Resetting limits for all columns.

Using a subset of data restricted by parameter range¶

You can use only a subset of the particles in both particle plots and Rendered plots, according to restrictions on any or all of the data columns (for example, using only particles with $\rho > 10$, in the 3D box $x,y,z \in [-0.1, 0.1]$). Whilst this has always been possible by selecting, colouring and/or hiding particles in Interactive mode (see Rendering using only a subset of the particles), the difference here is that the selection is based, for each timestep, strictly on the parameter range, rather than being a selection based on particle identity. This means that the parameter range is also saved to the splash.limits (i.e., by (s)aving your settings) and is shown when splash launches via lines such as:

>> current range restrictions set:

 (  1.693E-01 < x <  1.820E-01 )
 (  2.205E-01 < y <  2.265E-01 )
 (  7.580E-06 < density <  2.989E-05 )

>> only particles within this range will be plotted
   and/or used in interpolation routines

or more usually:

>> no current parameter range restrictions set

Parameter range restrictions can be set either manually via the (l)imits menu (option 7) or interactively by selecting a region in the plot and pressing ‘x’, ‘y’ or ‘r’ to restrict using the $x$, $y$ or both $x$ and $y$ limits of the selected area respectively (pressing ‘R’ instead removes all currently set restrictions). Another way of setting manual range restrictions is simply to edit the splash.limits file directly (this simply contains the min and max limits for each column, followed optionally by a third and fourth column specifying, respectively, the min and max of the range restriction).

Plotting only particles with $\rho > 10$, $u > 20$ and $-0.25 < x < 0.25$¶

Plotting a subset of the particles restricted by a parameter can be achieved by setting a parameter range restriction (which does not change between timesteps – see Using a subset of data restricted by parameter range), or alternatively by an interactive selection based on particle identity (see Rendering using only a subset of the particles).

(r)endering options¶

Changing the number of pixels in a rendered image¶

The number of pixels in a rendered image can be set manually using the (r)endering options, option 1 (or simply type r1 from the main menu). The number set is the number of pixels along the $x-$axis. The number of pixels along the $y-$axis is determined by the aspect ratio of the plot.

The number of pixels used in an image is, by default, automatically determined by the actual number of pixels available on the graphics device, which depends in turn on the size of the page (the page size can be set manually in the (p)age options – see Changing the size of the plotting surface). For vector (non-pixel) devices such as postscript, svg or pdf, the number of pixels is set to $1024/\textrm{n}$, where n is the number of panels across the page.

Changing the colour scheme¶

The colour scheme used for Rendered plots can be changed either by pressing m or M in Interactive mode to cycle through the available schemes or manually by using the change colour scheme option in the (r)endering options.

A demonstration of all the colour schemes can be also be invoked from this menu option. Setting the colour scheme to zero plots only the contours of the rendered quantity (assuming that plot contours is set to true). The colour schemes available are shown in Fig. 9.

splash colour schemes¶

User contributed colour schemes are eagerly invited (just send me either: a table of r,g,b colour indices [if you know them] or just an image of a colour bar you wish to reproduce and I will add it).

Plotting contours as well as the rendered image¶

Contours of either the rendered pixel array or of another (separate) quantity can be plotted on top of Rendered plots by setting the plot contours option from the (r)endering options. With this option set, an extra prompt will appear after the render prompt asking the user for a quantity to be contoured. The contoured quantity can also be set via the command line options (Command line options). If the rendered and contoured quantities are the same, further prompts appear which enable the limits for the contour plot to be set separately to the render plot. These limits are also saved separately in the splash.limits file when written.

Hint

To plot contours instead of the rendered image, use the change colour scheme option from the (r)endering options and choose colour scheme 0 (contours only).

Plotting contours instead of a rendered image¶

To plot contours instead of the rendered image, use the change colour scheme option from the (r)endering options and choose colour scheme 0 (contours only).

Changing the number of contour levels¶

The number of contour levels used whenever contours are drawn can be set via the change number of contours option in the (r)endering options. The contour levels can also be manually specified (see Setting the contour levels manually).

Setting the contour levels manually¶

As of v1.15.0, contour levels can be set manually by creating a file called splash.contours in the current directory (or prefix.contours if the splash -p prefix is specified on the command line). This file should contain one contour level per line, optionally with a label for each contour, e.g.

e-2  level 1
e-1  level 2
1    my really great contour
0    hi mum

Adding numeric labels to contours¶

An option to write numeric labels on contours appears as part of the change number of contours option in the (r)endering options.

Adding arbitrary contour labels¶

Contours can also be labelled manually by creating a splash.contours file. See Setting the contour levels manually.

Turning the colour bar off/ moving the colour bar label¶

The colour bar can be turned on or off and the style chosen (e.g. horizontal vs vertical) and for the vertical bar, the label moved closer to the bar itself, via the colour bar options option in the (r)endering options.

To change the text in the colour bar label, see Changing the label used for 3D projection plots.

Changing the style of the colour bar¶

The colour bar style (i.e., vertical vs. horizontal, plot-hugging vs. non plot-hugging, one-sided vs. two-sided, floating vs. fixed) can be changed via the colour bar options option in the (r)endering options. If you want a different style implemented, email me!

Using a horizontal colour bar¶

An option to use a horizontal colour bar instead of the default vertical arrangement is given in the colour bar options option in the (r)endering options.

Using ‘plot-hugging’ colour bars¶

See Changing the style of the colour bar.

Using floating/inset colour bars¶

See Changing the style of the colour bar.

Plotting ticks on only one side of the colour bar¶

See Changing the style of the colour bar.

Changing the text in the colour bar label¶

See Changing the label used for 3D projection plots.

Using coloured particles instead of rendering to pixels¶

As a simpler alternative to interpolating to a pixel array, particles can simply be coloured according to the value of a particular quantity by setting the use particle colours not pixels option in the (r)endering options. With this option set, Rendered plots are simply plotted by colouring the particles according to the rendered field. This is somewhat cruder but can be a good indication of where individual particles might be affecting results.

Danger

Any colouring of the particles set in Interactive mode will be overwritten by use of this option.

Using normalised interpolations¶

A normalised interpolation to pixels can be used by setting the normalise interpolations option from the (r)endering options. In general this leads to smoother rendering but also means that edges and surfaces appear more prominently (and a bit strange). The rule-of-thumb I suggest is to use this option when there are no free surfaces in the simulation.

Warning

Normalising a 3D column density rendering means plotting the quantity $\int \rho {\rm d}z / \int {\rm d}z$, which is a bit meaningless. More useful is to turn on Density weighted interpolation, which gives a mass-weighted line of sight average. For example, plotting temperature, this would give $\int \rho T {\rm d}z / \int \rho {\rm d}z$ which is more meaningful.

Speeding up the rendering on 3D column integrated plots¶

Interpolation on 3D column integrated plots can be made faster by setting the use accelerated rendering option in the (r)endering options. The reason this is an option is that it makes a small approximation by assuming that each particle lies exactly in the centre of a pixel. In general this works very well but is not set by default because it can produce funny looking results when the particles are aligned on a regular grid (e.g. as is often the case in initial conditions). Typical speed-ups range from $\times 2$ up to $\times 4$, so it is highly recommended for interactive work.

Density weighted interpolation¶

Density weighted interpolation (where a quantity is plotted times $\rho$) can be turned on in the (r)endering options.

Selecting and rendering only a subset of the particles¶

An example of how to render using only a selected subset of the particles was given in Rendering using only a subset of the particles.

Changing the label used for 3D projection plots¶

The labelling scheme used to determine the colour bar label can be changed via the customize label on projection plots option in the (r)endering options. Information specific to the quantity being rendered can be incorporated via format codes as follows:

Example format strings:
 \(2268) %l d%z %uz       : this is the default format "\int rho [g/cm^3] dz [cm]"
  column %l               : would print "column density" for density
 surface %l               : would print "surface density"
 %l integrated through %z : would print "density integrated through z"

Format codes:
%l  : label for rendered quantity
%z  : label for 'z'
%uz : units label for z (only if physical units applied)

Changing “column density” to “surface density” on 3D plots¶

See Changing the label used for 3D projection plots.

Changing the interpolation kernel¶

The kernel used for the interpolations is by default the M$_{4}$ cubic B-spline, which has been standard in SPH calculations since the mid-1980’s. Other kernels can be selected via the change kernel option in the (r)endering options. The kernel can also be changed by setting the SPLASH_KERNEL environment variable to either the kernel name as listed in the (r)endering options, or something sensible resembling it. At present only a few kernels are implemented, with cubic , quartic and quintic referring to the M$_{4}$, M$_{5}$ and M$_{6}$ B-splines with support of 2h and 3h, respectively. See [Price12] for more details.

(v)ector plot options¶

Changing the number of arrows on vector plots¶

See Changing the number of pixels in vector plots.

Changing the number of pixels in vector plots¶

The number of pixels used on Vector plots can be changed via the change number of pixels option in the (v)ector plot options. This controls the number and average size of the arrows which appear (i.e., one arrow is plotted at the centre of each pixel).

Changing the size of arrows on vector plots¶

The size of the arrows on Vector plots is proportional to the magnitude of the vector quantity at that pixel, where the maximum size is set from the maximum plot limit for the x, y and z components of the vector quantity being plotted such that the longest arrow fills one pixel. These limits can be changed manually via the (l)imits menu options. Where these limits are nowhere near the actual values of the vector field, arrows can appear either very big (just a line across the screen) or extremely small (appearing as just dots). Pressing w in Interactive mode automatically adjusts the arrows to sensible proportions (this is the equivalent of pressing a for non-vector quantities). Alternatively pressing v (to decrease) or V (to increase) can be used to adjust the arrow lengths (the change can be multiplied by 10 or more by first pressing z one or more times before pressing v or V).

Plotting vector arrows in white instead of black or vice-versa¶

Vector arrows are by default plotted using the current foreground colour index (i.e., as used for plotting the axes). To plot in the background colour index instead set the use background colour for arrows option in the (v)ector plot options.

Turning off the legend for vector plots¶

The legend which appears on Vector plots can be turned on or off via the vector plot legend settings option in the (v)ector plot options.

Moving the vector plot legend¶

The position of the Vector plots legend can be set either interactively by positioning the mouse and pressing H or manually via the vector plot legend settings option in the (v)ector plot options.

Plotting stream/fieldlines instead of arrows¶

To plot Vector plots that use stream/fieldlines instead of arrows, set the plot stream/field lines instead of arrows option in the (v)ector plot options. This option performs a simple integration of the interpolated vector field to get the stream function, the contours of which are then plotted

Hint

The number of contours can be changed via the change number of contours option in the (r)endering options. It is generally advantageous to use a larger number of pixels for the vector interpolation (See Changing the number of pixels in vector plots) to get smooth contours.

At present this option works quite well for smooth vector fields but can perform poorly for vector fields with strong gradients.

Turning arrow heads off for vector plots¶

Vector plots can be plotted using arrows without heads using the turn arrow heads on/off option in the (v)ector plot options.

Hiding vector arrows where there are no SPH particles¶

On Rendered plots often arrows can appear where there are apparently no SPH particles because the interpolation is performed to all pixels within $2h$ of an SPH particle. Such arrows in regions of few or no particles can be hidden using the hide arrows where there are no particles option in the (v)ector plot options. A threshold number of particles for each pixel can be specified, below which no arrow will be plotted on that pixel.

Plotting a vector plot in a cross section slice¶

Vector plots are either in a Cross section or are column integrated projections depending on the setting of the switch between cross section/projection option in the (x) cross section/3D plotting options. Setting this to cross section and plotting Vector plots produces a vector plot in a cross section slice.

Making all arrow the same length (i.e., showing direction only, not magnitude)¶

An option to plot all vector arrows of the same length (instead of the default option where the length of the arrow is proportional to the vector magnitude) can be set from the (v)ector plot options.

(x) cross section/3D plotting options¶

Plotting a cross section slice through 3D data¶

When plotting Rendered plots of 3D data, the default option is to plot a column-integrated plot. To change this to a Cross section, use option 1) in the (x) cross section/3D plotting options (switch between cross section/projection). See Basic splash usage for examples of how this works. An oblique Cross section can be set interactively using the x key, see Taking an oblique cross section interactively which works by setting a combination of rotation and a Cross section position.

Plotting a cross section line through 2D data¶

In 2D, setting the switch between cross section/projection option in the (x) cross section/3D plotting options to Cross section means that Rendered plots are in fact a 1D Cross section (i.e., a line) through 2D data. The position of the line is completely arbitrary (i.e., can be set for oblique cross sections as well as straight lines) and is set interactively after the usual $y-$ and $x-$ axis prompts.

Rotating the particles¶

An angle of rotation about may be set each axis may be set in the (x) cross section/3D plotting options using the rotation on/off/settings option or interactively (press h in Interactive mode to see the exact keystrokes). The position of the origin about which particles are rotated can be set from the rotation on/off/settings option in the (x) cross section/3D plotting options. Rotated axes or boxes can be plotted using the set axes for rotated/3D plots option in the same menu.

Rotations are performed in the order $z-y-x$. This means that the $y-$ rotation angle is an angle about the new $y-$axis, defined by the $z$ rotation and similarly for the $x-$ rotation. If you think about it long enough, it makes sense. If in doubt, do it interactively and set the angles in the order $z-y-x$.

Setting the origin about which particles are rotated¶

The origin about which particles are rotated and relative to which the radius is calculated when the calculate extra quantities option is set in the (d)ata options can be changed via the rotation on/off/settings option in the (x) cross section/3D plotting options.

Adding 3D perspective¶

3D perspective can be turned on via the 3D perspective on/off option in the (x) cross section/3D plotting options. Prompts for setting the perspective position then appear after the usual prompts for y and x axes, Rendered plots and Vector plots, i.e., something like the following:

Please enter your selection now (y axis or option):2
(x axis) (default=1):
 (render) (0=none) ([0:20], default=0):
 (vector plot) (0=none, 7=B, 10=v, 17=J) ([0:17], default=0):
 enter z coordinate of observer (default=1.800):
 enter distance between observer and projection screen ([0.000:], default=0.1800):
 Graphics device/type (? to see list, default /xwin):

3D perspective is defined by two parameters: a distance to the observer $zobs$ and a distance between the observer and a screen placed in front of the observer, $dscreen$. The transformation from usual $x$ and $y$ to screen $x'$ and $y'$ is then given by

\[\begin{split}\begin{aligned} x' & = & x*dscreen/(zobs-z), \nonumber \\ y' & = & y*dscreen/(zobs-z).\end{aligned}\end{split}\]

This means that objects at the screen distance will have unit magnification, objects closer than the screen will appear larger (points diverge) and objects further away will appear smaller (points converge). The situation could be beautifully illustrated if I could be bothered drawing a figure. I have found reasonable results with something like a $1/10$ reduction at the typical distance of the object (i.e., observer is placed at a distance of $10\times$ object size with distance to screen of $1\times$ object size). splash sets this as default using the z plot limit as the ‘object size’.

Hint

The position of the 3D observer in $z$ can also be changed in Interactive mode using u or U (to move ’up’) and d or D (to move down).

Using 3D surface rendering¶

3D surface rendering (turned on using the 3D surface rendering on/off option in the (x) cross section/3D plotting options) performs a ray-trace through the particle data, thus visualising the “last scattering surface”. When set, the user is prompted for an “optical depth” before plotting which determines the position of the surface. Only applies to 3D data. When set with cross-section (instead of projection), particles at or below the z value of the slice are used.

For examples of the 3D surface rendering in splash , have a look at my movies of neutron star mergers:

http://users.monash.edu.au/~dprice/research/nsmag.

Plotting 3D box / 3D axes¶

Rotated axes or boxes can be plotted using the set axes for rotated/3D plots option in the (x) cross section/3D plotting options.

Setting up animation sequences¶

Animation sequences can be set via the set animation sequence option in the (x) cross section/3D plotting options. At present the possible sequences that can be added are:

: steady zoom on x and y axes
: steady rotation
: steady change of limits (e.g. for colour bar)
: steady movement of 3D observer
: sequence of cross section slices through a 3D box
: steady change of opacity for 3D surface plots

Up to one sequence of each type can be added (i.e., up to 6 in total) with different start and end points (specified in terms of dump file number), with the additional possibility of inserting extra frames between dump files (e.g. to plot a sequence of frames consisting of a changing view of the same dump file).

Animation sequences can also be set using e in Interactive mode. To set a sequence interactively first adjust the plot settings to correspond to the start of the sequence (pressing s to save if this is done in Interactive mode). Then in Interactive mode move to the dump file you want to be the end-point and also adjust the plot settings to correspond to the end-point of your desired sequence (i.e., adjust the colour bar limits and/or adjust the rotation angle and/or the x/y limits and/or the 3D observer position and/or the opacity). Then, rather than pressing s (which would make these become the default plot settings) press e instead, saving these settings as the end-point of the desired animation sequence. This can be done multiple times to set multiple sequences.

Animation sequences set up in this manner are saved to a file called splash.anim either when prompted (if setting sequences non-interactively) or by pressing ’S’ from the main menu which then saves both the splash.limits and splash.anim files in addition to the usual splash.defaults file.

Note: Animation sequences act on a ‘per page’ basis rather than simply ‘per frame’. This means that you can produce a multi-panelled movie (e.g.) showing the evolution of different runs side by side, with the same animation sequence applied to each.

Plotting a sequence of frames rotating a data set through 360 degrees¶

This can be achieved by setting an animation sequence with a steady change of rotation angle. See Setting up animation sequences.

Plotting a ‘fly-around’ of 3D data¶

This can be achieved by setting an animation sequence with a steady change of rotation angle. See Setting up animation sequences.

Plotting a flythru of 3D data¶

A sequence of Cross section slices progressively deeper into a 3D box or alternatively a steady movement of the 3D observer (on projection plots) can be plotted by setting up an animation sequence. See Setting up animation sequences.

Adding a steady zoom sequence to a movie¶

A steady change of $x-$ and $y-$ limits can be added by setting up an animation sequence. See Setting up animation sequences.

Adding a steady change of colour bar limits¶

A steady change of limits on the colour bar over one or more dump files for a movie can be implemented by setting up an animation sequence. See Setting up animation sequences for details.

Adding steady movement of the 3D observer¶

The position of the 3D observer can be steadily changed over several dump files (or several frames produced of the same dump file) by setting up an animation sequence. See Setting up animation sequences for details.

Miscellaneous other useful things¶

My attempt at in-built help¶

The My attempt at in-built help does nothing particularly useful apart from tell you about menu shortcuts (see Keyboard shortcuts to menu options). It seemed like a good idea at the time…

Exiting splash¶

(q)uit, unsurprisingly, quits. Typing a number greater than the number of data columns also exits the program (e.g. I often simply type 99 to exit).

Advanced plotting examples¶

Rendered plot of star formation data¶

This is an example using data provided by Paul Clark. The data is from an SPH simulation of star formation in sphNG format. We read the dump file as follows:

dprice/dustfrag> ssplash omuk162

after which we get output along the lines of:

...
 reading single dumpfile
>>>>>>>>>>>>>>>>>>>>>>>>>> omuk162 <<<<<<<<<<<<<<<<<<<<<<<<<<
double precision dump
File ID: FHydroRTMHD1
 npart =  11744854
...

and arrive at the main menu:

 You may choose from a delectable sample of plots
-------------------------------------------------------
  1) x                                7) v\dx
  2) y                                8) v\dy
  3) z                                9) v\dz
  4) particle mass        10) u
  5) h                              11) grad h
  6) density
-------------------------------------------------------
 12) multiplot [  4 ]      m) set multiplot
-------------------------------------------------------
 d(ata) p(age) o(pts) l(imits) le(g)end h(elp)
 r(ender) v(ector) x(sec/rotate) s,S(ave) q(uit)
-------------------------------------------------------
Please enter your selection now (y axis or option):

here we want to plot a rendered plot of column density (density is in column 6), so we type 2 for column 2 (y) as the y axis, 1 for column 1 (x) as the x-axis and at the render prompt 6, for density, ie:

Please enter your selection now (y axis or option):2
(x axis) (default=1):
 (render) (0=none) ([0:11], default=0):6
 (vector plot) (0=none, 7=v) ([0:7], default=0):0
 Graphics device/type (? to see list, default /xw): /xw

producing the plot shown in Fig. 10 – somewhat black!

First stage in the star formation figure tutorial: a simple render plot of density¶

The main thing to note is the limits on the colour bar (extending from $0$ to $10^{7}$ on a linear scale) which is the main source of all the blackness. Moving the cursor over the colour bar and pressing l for log produces Fig. 11 — a vast improvement!

With a log axis after having placed cursor over colour bar and pressed l¶

For this visualisation we will eventually want the data in physical units rather than code units. For the sphNG read these units are already specified in the read_data routine, so all we have to do is turn physical units on. Pressing q from interactive mode (that is, with the cursor in the plot window) returns us to the main menu.

Physical units are turned on from the (d)ata options, as follows:

Please enter your selection now (y axis or option):d
----------------- data read options -------------------
 0) exit
 1) read new data /re-read data
 2) change number of timesteps used        (     1 )
 3) plot selected steps only               (  OFF )
 4) buffering of data on/off               (  OFF )
 5) turn calculate extra quantities on/off (  OFF )
 6) use physical units                     (  OFF )
 7) change physical unit settings
enter option ([0:7], default=0):6
 current settings for conversion to physical units are:
x [cm] = x x  1.000E+17
y [cm] = y x  1.000E+17
z [cm] = z x  1.000E+17
particle mass [g] = particle mass x  1.991E+33
h [cm] = h x  1.000E+17
density [g/cm\u3\d] = density x  1.991E-18
v\dx [cm/s] = v\dx x  3.645E+04
v\dy [cm/s] = v\dy x  3.645E+04
v\dz [cm/s] = v\dz x  3.645E+04
u [erg/g] = u x  1.328E+09
grad h = grad h x  1.000E+00
time = time x 1.69E+00
Use physical units? (default=yes):

returning us to the main menu with labels changed as follows:

You may choose from a delectable sample of plots
-------------------------------------------------------
  1) x [cm]                7) v\dx [cm/s]
  2) y [cm]                8) v\dy [cm/s]
  3) z [cm]                9) v\dz [cm/s]
  4) particle mass [g]    10) u [erg/g]
  5) h [cm]               11) grad h
  6) log density [g/cm\u3
-------------------------------------------------------
 12) multiplot [  4 ]      m) set multiplot
-------------------------------------------------------
 d(ata) p(age) o(pts) l(imits) le(g)end h(elp)
 r(ender) v(ector) x(sec/rotate) s,S(ave) q(uit)
-------------------------------------------------------
Please enter your selection now (y axis or option):

at this stage we will save the current settings to file by pressing s from the main menu.

Please enter your selection now (y axis or option):s
 default options saved to file splash.defaults

Actually we would prefer the column labels in AU, but we will come to that later. Replotting the same plot (that is 2, 1, 6, 0, /xw from the main menu) plots the same plot we had before, but with the axes in physical units. Zooming in (using the mouse) on the region of interest and adapting the colour bar limits by moving the mouse over the colour bar and pressing a produces the plot shown in Fig. 12.

Second stage in the star formation figure tutorial: having applied physical units, zooming in and pressing a on the colour bar¶

For this kind of plot, the Bate colour scheme looks better – pressing m with the mouse in the plot window changes the colour scheme, producing the plot shown in Fig. 13.

having changed the colour scheme¶

Pressing s in interactive mode (that is, with the mouse in the plot window) saves the current zoom and colour bar settings (but not to disk until you also press S from the main menu). Pressing q from interactive mode returns to the main menu.

Next we want to turn on the plotting of sink particles (all particle types other than gas are turned off by default). This is done in the particle plot (o)ptions as follows:

 Please enter your selection now (y axis or option):o
------------- particle plot options -------------------
 0) exit
 1) turn on/off particles by type       ( ON, OFF, OFF, OFF )
 2) change graph markers for each type  (  1,  4, 17,  1 )
 3) set colour for each particle type   ( -1, -1, -1, -1 )
 4) plot line joining particles         ( OFF )
 5) plot smoothing circles              (   0 )
 6) use fast particle plotting          ( ON  )
 7) change coordinate systems           (  1 )
 8) plot exact solution                 (  0 )
 9) exact solution plot options
enter option ([0:9], default=0):1
 Plot gas particles? (default=yes):
 Plot ghost particles? (default=no):
 Plot sink particles? (default=no):y
 >> Plot sink particles on top of rendered plots? (default=no):y
 Plot unknown/dead particles? (default=no):

Repeating our previous plot (i.e., 2, 1, 6, 0, /xw) produces the plot shown in Fig. 14.

Third stage in the star formation figure tutorial: having turned sink particle plotting on¶

replacing the axes with a scale¶

The axes in [cm] are kind of ugly, so we could either change this to a sensible unit or plot a scale instead. We will do the latter. The axes can be turned off in the (p)age options, as follows:

 Please enter your selection now (y axis or option):p
---------------- page setup options -------------------
...
 2) axes options                      ( 0)
...
enter option ([0:8], default=0):2
  -4 : draw box and major tick marks only;
 -3 : draw box and tick marks (major and minor) only;
 -2 : draw no box, axes or labels;
 -1 : draw box only;
  0 : draw box and label it with coordinates;
  1 : same as AXIS=0, but also draw the coordinate axes (X=0, Y=0);
  2 : same as AXIS=1, but also draw grid lines at major increments of the coordinates;
 10 : draw box and label X-axis logarithmically;
 20 : draw box and label Y-axis logarithmically;
 30 : draw box and label both axes logarithmically.
enter axis option ([-4:30], default=0):-2
   axis =  -2

The option to plot a scale of a particular length is also to be found in the le(g)end and title options. We will choose to plot a scale of length 0.1 pc.

Please enter your selection now (y axis or option):g
---------------- legend and title options -------------------

 To set the plot titles, create a file called
  'splash.titles' in the working directory, with one title per line

 0) exit
 1) time legend on/off/settings                ( ON   0.87  1.87  0.00 "t=")
 2) titles on/off/settings                     ( ON   0.20 -0.92  0.00)
 3) legend for multiple steps per page on/off  ( OFF )
 4) plot scale on co-ordinate plots            ( OFF )
 5) legend only on nth panel/first row/column  (  0 )
Enter option ([0:5], default=0):4
 Plot scale on co-ordinate plots? (default=no):y
 Enter length of scale in the current x,y,z units (default=1.000):3.0856e15
 Enter text to appear below scale (e.g. '10 AU') (default=1 unit): 0.1 pc
 Enter horizontal position as fraction of viewport ([0.000:1.000], default=0.5000):
 Enter vertical position in character heights above bottom (default=1.000):

Note that because the x axis units were already in cm, we simply entered the value for 0.1pc in these units. Before plotting again, we should save what we have done so far to disk: Pressing S from the main menu saves both the current plot settings and the plot limits to disk:

Please enter your selection now (y axis or option):S
 default options saved to file splash.defaults
 saving plot limits to file splash.limits

Plotting our figure again (2-1-6-0-/xw) produces the plot shown in Fig. 14

Nearly there…! To add the finishing touches we want to increase the number of pixels substantially. This is done in the (r)endering options, option 1, for which we can use the shortcut r1:

Please enter your selection now (y axis or option):r1
----------------- rendering options -------------------
enter number of pixels along x axis ([1:10000], default=200):1000

then, to plot the figure to file instead of the screen, we simply choose a different device at the prompt:

Please enter your selection now (y axis or option):2
(x axis) (default=1):
 (render) (0=none) ([0:11], default=6):
 (vector plot) (0=none, 7=v) ([0:7], default=0):
 Graphics device/type (? to see list, default /xw): starpartfinal.gif/gif

producing our final finished Fig. 16.

Finished star formation plot¶

Pressing S from the main menu saves all of the settings and plot limits to disk, so invoking splash again will produce the same plot. To produce the same plot on a sequence of dumps, simply put more than one file on the command line and plot to a non-interactive device (see Producing a sequence of plots for a movie). Use the postscript devices /ps or /cps (for colour) to make figures suitable for inclusion in a paper.

Other things you may want to do with this plot include:

Turn the time legend off. See Turning off/moving the time legend.
Change the colour of sink particles. See Plotting each particle type in a different colour.
Change the foreground/background colour of the page. See Changing the foreground and background colours.

Multi-panelled figure¶

The following is an example plot taken from [PB07]. Here I will plot a sequence of plots tiled on the same page, so that columns correspond to dumps taken from different runs at the same time and rows correspond to an evolutionary sequence from a given run. The plot uses sphNG data which contains sink particles, so I also want these to appear on the plots and be plotted in white. Basically I want the plots to be plotted such that as much of the plot is taken up by data and very little by axes and the like but still conveying all of the necessary information.

We proceed as follows: Firstly, each different run (corresponding in this case to a series of runs with different magnetic field strength) are in different subdirectories with names like mbossbod_f10.0/, mbossbod_f5.0/, etc. which all contain a sequence of dump files with names like mbos001, mbos002 etc. To begin the plot, I start by creating a new, empty subdirectory so that the splash.defaults and splash.limits files created by pressing S from the main menu will be in this directory such that running splash from that directory always produces this plot. So:

dprice% mkdir plot1
dprice% cd plot1

then having decided which dump files from each run to use, I create a text file listing these filenames (with the full relative pathname) in the order in which I will plot them. For example, calling this file (arbitrarily) filelistplot, the contents should be something like the following:

dprice% more filelistplot
../mbossbod_f20.0/mbos259
../mbossbod_f20.0/mbos263
../mbossbod_f20.0/mbos268
../mbossbod_f20.0/mbos275
../mbossbod_f20.0/mbos294
../mbossbod_f10.0/mbos259
../mbossbod_f10.0/mbos263
../mbossbod_f10.0/mbos268
../mbossbod_f10.0/mbos275
../mbossbod_f10.0/mbos294
../mbossbod_f7.5/mbos259
../mbossbod_f7.5/mbos263
../mbossbod_f7.5/mbos268
...

Then invoke splash (ssplash for sphNG) with these filenames on the command line:

ssplash `cat filelistplot`

after which the first dump file should be read, indicated by output along the lines of:

 reading single dumpfile
>>>>>>>>>>>>>>>>>>>>>>>>>> ../mbossbod_f20.0/mbos259 <<<<<<<<<<<<<<<<<<<<<<<<<<
double precision dump
File ID: SHydroRTMHD1
 npart =  491567
...

An alternative method is to rename the filelistplot file splash.filenames, from which the filenames will be read if there are none specified on the command line (this feature was implemented as a workaround for a limit to the number of command line arguments on the some compilers).

The first stage is to get a plot of a single panel looking good. So, from the main menu, we will plot a simple rendering of density and adjust the plot limits until we are happy:

 You may choose from a delectable sample of plots
-------------------------------------------------------
  1) x                     6) density
  2) y                     7) B\dx
  3) z                     8) B\dy
  4) particle mass         9) B\dz
  5) h
-------------------------------------------------------
 10) multiplot [  4 ]      m) set multiplot
-------------------------------------------------------
 d(ata) p(age) o(pts) l(imits) le(g)end h(elp)
 r(ender) v(ector) x(sec/rotate) s,S(ave) q(uit)
-------------------------------------------------------
Please enter your selection now (y axis or option):2
 (x axis) (default=1):
 (render) (0=none) ([0:9], default=0):6
 (vector plot) (0=none, 7=B) ([0:7], default=0):
  Graphics device/type (? to see list, default /xw): /xw

which should produce the plot shown in Fig. 17. Not much can be seen at first – just a few white dots. This is mainly a result of the density axis (i.e., the colour bar) not being logged.

First stage in the multi-panelled figure tutorial: a simple render plot of density¶

Moving the cursor over the colour bar and pressing l results in the plot shown in Fig. 18

and with a log axis after having placed cursor over colour bar and pressed l¶

Before we proceed any further, we will first change the axes to be in physical units rather than code units. Pressing q in the plot window to exit interactive mode and return to the main menu, and from the (d)ata options, turn the use physical units option on:

Please enter your selection now (y axis or option):d
----------------- data read options -------------------
 0) exit
 1) read new data /re-read data
 2) change number of timesteps used        (     1 )
 3) plot selected steps only               (  OFF )
 4) buffering of data on/off               (  OFF )
 5) turn calculate extra quantities on/off (  OFF )
 6) use physical units                     (  OFF )
 7) change physical unit settings
enter option ([0:7], default=0):6
 current settings for conversion to physical units are:
x [cm] = x x  1.000E+16
y [cm] = y x  1.000E+16
z [cm] = z x  1.000E+16
particle mass [g] = particle mass x  1.991E+33
h [cm] = h x  1.000E+16
density [g/cm\u3\d] = density x  1.991E-15
B\dx [G] = B\dx x  1.000E+00
B\dy [G] = B\dy x  1.000E+00
B\dz [G] = B\dz x  1.000E+00
time = time x 1.13E-01
Use physical units? (default=yes):yes

The default transformations to physical units are in this case set in the data read. However it would be nicer in this case to set the x and y axis units to AU (Astronomical Units), rather than cm. From the (d)ata options we proceed as follows:

enter option ([0:7], default=0):7
 enter column to change units (-2=reset all,-1=quit,0=time) ([-2:9], default=-1):1
 enter x [cm] units (new=old*units) (default=0.1000E+17):668.3893
 enter label amendment (default=[cm]): [AU]
 Apply these units to all coordinates and h? (default=yes):
 Enter unit for 'z' in 3D column integrated plots (default=0.1000E+17):
 Enter label for z integration unit (e.g. [cm]) (default=[cm]):

enter column to change units (-2=reset all,-1=quit,0=time) ([-2:9], default=-1):

save units to file? (default=yes):
 saving plot limits to file splash.units

where in the above I set the multiplicative factor such that the x axis will be in AU and correspondingly changed the units label to [AU] (note the preceding space). I was also prompted to change the unit for ’z integration’ – this is the length unit added when integrating a quantity through z. Leaving this in cm means that, even though the coordinate axes are in AU, the density (in g/cm$^{3}$) is integrated through z in cm, giving column density in g/cm$^{2}$ (as opposed to g /cm:math:^{3} AU).

To save what we have done so far, press s from the main menu to save the current settings to the splash.defaults file:

Please enter your selection now (y axis or option):s
default options saved to file splash.defaults

Having turned physical units on, we replot the same plot (i.e., answering 2, 1, 6, 0, /xw to the prompts, as previously). First of all we find simply a white screen. This is a result of the colour bar axis now being wrong. Moving the mouse over the colour bar and pressing a (to adapt) results in the plot shown in Fig. 19. The plot looks basically identical to the previous plot, except that the axes are now in physical units (x and y are in AU and column density is in g/cm$^{2}$).

Second stage in the multi-panelled figure tutorial: having changed the axes into physical units¶

Next, we zoom in to the central region of interest using the mouse – selecting a region and clicking to zoom in. Pressing o centres the plot on the origin and as we zoom in it we also press a over the colour bar to readjust the colour bar limits to the max/min on the zoomed-in plot. Finishing with the adjustments (and pressing s in the plot window to save the current settings) results in the plot shown in Fig. 20.

zooming in and adjusting the colour bar¶

Surface rendering¶

Here I will give an example of how to use the 3D surface rendering feature starting with a dump file kindly supplied by Giuseppe Lodato from an SPH simulation of a warped accretion disc. First we read the file (in sphNG format, so we use ssplash):

dprice$ ssplash warp001

after which we reach the main menu:

 You may choose from a delectable sample of plots
-------------------------------------------------------
  1) x                     6) density
  2) y                     7) v\dx
  3) z                     8) v\dy
  4) particle mass         9) v\dz
  5) h
-------------------------------------------------------
 10) multiplot [  4 ]      m) set multiplot
-------------------------------------------------------
 d(ata) p(age) o(pts) l(imits) le(g)end h(elp)
 r(ender) v(ector) x(sec/rotate) s,S(ave) q(uit)
-------------------------------------------------------
Please enter your selection now (y axis or option):

Firstly we want to plot just a simple render plot of density. Thus we choose:

 Please enter your selection now (y axis or option):2
(x axis) (default=1):
 (render) (0=none) ([0:9], default=0):6
 (vector plot) (0=none, 7=v) ([0:7], default=0):
 Graphics device/type (? to see list, default /xwin): /xw

producing the plot shown in Fig. 21 (I have used /png instead of /xw to produce the figures for the userguide).

First stage in the surface rendering tutorial: a simple render plot of density¶

Moving the cursor over the colour bar and pressing l to log the colour bar axis produces Fig. 22.

with a log axis after having placed cursor over colour bar and pressed l¶

The next step is to adjust the viewing angle. Pressing h in the plot window brings up the list of keystrokes which can be used to change the angle. Here we want to add a rotation about the $x-$ axis, so we press { three times to change the x angle by -90 degrees and then press [ once to increment the angle by a further -15 degrees. The splash output in the terminal reads, amongst other things:

rotating particles about z by   0.00
rotating particles about y by   0.00
rotating particles about x by 255.00

Then we obtain Fig. 23.

Second stage in the surface rendering tutorial: after adjusting the rotation angle¶

Next, we need to turn the 3D surface rendering on. This cannot be done in interactive mode so we need to exit – pressing s first to save what we have done so far, then q to quit interactive mode. Then, back at the splash main menu, we type x4 for the (x) cross section/3D plotting options, option 4 which is “3D surface rendering on/off” with prompts appearing as follows:

Please enter your selection now (y axis or option):x4
---------- cross section / 3D plotting options --------
Use 3D opacity rendering? (default=yes):y

Now we replot the original plot with the new settings as follows:

Please enter your selection now (y axis or option):2
(x axis) (default=1):
 (render) (0=none) ([0:9], default=6):
 (vector plot) (0=none, 7=v) ([0:7], default=0):
 enter z coordinate of observer (default=53.58):
 enter distance between observer and projection screen ([0.000:], default=5.358):
 using current h and pmass limits to calculate kappa (cross section/unit mass)
 min h =  0.1197254  min particle mass =  3.812551E-11
 [ kappa = pi*h_min**2/(particle_mass*n_smoothing_lengths) ]
enter approximate surface depth (number of smoothing lengths): ([0.000:], default=2.000):
 kappa (particle cross section per unit mass) =  1.2369025E+9
 Graphics device/type (? to see list, default /xwin):

Note that several new prompts appear – for the moment I have just used the default answers by pressing return. The first result is rather frightening : just a black image with a black colour bar! This is because the limits we set for column density are several orders of magnitude away from the limits on density. Moving the cursor over the colour bar and pressing a to adapt the limits produces the plot shown in Fig. 24.

With 3D surface rendering turned on (which also turns on 3D perspective) and having adjusted the colour bar limits¶

Note that the plot suddenly appears much smaller – this is a consequence of the 3D perspective settings. Moving the cursor into the plot window and pressing a adapts the plot limits. After also clicking on the colour bar and adjusting the colour bar limits, we arrive at the plot shown in Fig. 25.

Third stage in the surface rendering tutorial: after adjusting the xy and colour bar limits interactively¶

Now that we are nearly there, to add the finishing touches we need to i) increase the number of pixels in the image and ii) turn the axes off, since they are no longer meaningful with 3D perspective set. The number of pixels can be increased by returning to the splash main menu (pressing s in interactive mode before doing so to save what we have done so far), then typing r1 for (r)endering options, option 1:

Please enter your selection now (y axis or option):r1
----------------- rendering options -------------------
enter number of pixels along x axis ([1:10000], default=200):1000

Next, we turn the axes off using the (p)age options:

Please enter your selection now (y axis or option):p2
---------------- page setup options -------------------
 -4 : draw box and major tick marks only;
 -3 : draw box and tick marks (major and minor) only;
 -2 : draw no box, axes or labels;
 -1 : draw box only;
  0 : draw box and label it with coordinates;
  1 : same as AXIS=0, but also draw the coordinate axes (X=0, Y=0);
  2 : same as AXIS=1, but also draw grid lines at major increments of the coordinates;
 10 : draw box and label X-axis logarithmically;
 20 : draw box and label Y-axis logarithmically;
 30 : draw box and label both axes logarithmically.
enter axis option ([-4:30], default=0):-2
  axis =  -2

Plotting the same plot again now results in the plot shown in Fig. 26.

and increasing the number of pixels and having turned the axes off¶

Finally we will also set the background colour to black, adjust the opacity and move the time legend. Notice that in Fig. 26 the surface looks blotchy. This is an indication that the surface is too shallow (that is we are only seeing particles on the very top). Thus we will adjust the opacity for a slightly deeper plot. We proceed as follows: Exiting interactive mode (pressing s then q in the plot window), we first set the foreground and background colours in the (p)age options:

Please enter your selection now (y axis or option):p8
---------------- page setup options -------------------
Enter background colour (by name, e.g. "black") (default=):black
 Enter foreground colour (by name, e.g. "white") (default=):white
 Do you want to plot axes and overlaid text in background colour (default is foreground) ? (default=no):

Now, replotting the same plot again, but this time adjusting the opacity at the prompt:

enter approximate surface depth (number of smoothing lengths): ([0.000:], default=2.000):200.0

Finally, moving the time legend by positioning the cursor and pressing ’G’ and zooming out slightly by pressing - once, we arrive at our finished figure (or movie frame) shown in Fig. 27. Pressing s in interactive mode saves the settings, then pressing q returns to the splash main menu. To (s)aving your settings to disk, press S from the main menu to save both the splash.defaults file and the splash.limits file.

Finished surface-rendered plot¶

To create a sequence of images with these settings, then simply invoke splash again with multiple files:

ssplash warp???

then plotting the same plot as previously to a non-interactive device will cycle through all dump files producing a sequence of plots with names like splash_0000.png, splash_0001.png etc. These can be easily converted into an animation.

Using splash to plot energy vs time plots¶

splash can be used for all kinds of data. For example I often use it to plot the contents of the .ev file my SPH code dumps monitoring quantities like energy and angular momentum at every timestep. A shortcut way of setting options appropriate to reading such files (e.g. plotting lines instead of dots, plotting all files on the same page) is implemented by adding the “-ev” option to the command line: e.g.

splash -ev file1.ev file2.ev file3.ev

also, using the -ev option on the command line means that any modification to the preset options /limits are saved to files called evsplash.defaults and evsplash.limits instead of the usual splash.defaults and splash.limits. This means the defaults for this type of plot are saved separately to those for “normal” plots of SPH data.

For other command line options, see Command line options.

Plotting azimuthally-averaged disc surface density and Toomre Q parameter¶

For analysis of accretion disc simulations, it is useful to make azimuthally averaged plots of the disc properties such as the surface density and, for self-gravitating discs, the Toomre Q parameter. Extra columns appear to plot both of these quantities when the simulation is 3D and the coordinate system is changed to cylindrical or spherical co-ordinates (in the particle plot (o)ptions – see Plotting in different coordinate systems (e.g. cylindrical coordinates)). For the Toomre Q parameter to appear it is also necessary to have read the thermal energy from the dump file. For example, having read a dump file, change the coordinate system to cylindricals:

Please enter your selection now (y axis or option):o7
------------- particle plot options -------------------
 0) reset (= 1)
 1) cartesian   x,y,z
 2) cylindrical r,phi,z
 3) spherical   r,phi,theta
 4) toroidal    r,theta,phi
Enter coordinate system to plot in: ([0:4], default=1):2

then extra columns appear in the menu:

 You may choose from a delectable sample of plots
-------------------------------------------------------
  1) r                    13) u
  2) phi                  14) grad h
  3) z                    15) grad soft
 ...                         ...
 11) v\dphi               23) Surface density
 12) v\dz                 24) Toomre Q parameter
-------------------------------------------------------

Then (in this example), select column 23 to plot surface density,

Please enter your selection now (y axis or option):23
 setting x axis to r for surface density plot

…and the plot will appear - an example surface density plot is shown in Fig. 28.

Plot of azimuthally averaged surface density in a 3D accretion disk simulation¶

Azimuthally averaged quantities are calculated by binning the particles into a fixed number of annuli in radius. The mean surface density is calculated using

\[\Sigma(r_{ann}) = \frac{M_{ann}}{\pi [(r_{ann} + 0.5\Delta r)^{2} - (r_{ann} - 0.5\Delta r)^{2}]},\]

that is, the total mass in the annulus (sum of the particle masses) divided by its area, where $r_{ann}$ is the radius (cylindrical or spherical) of the annulus. The Toomre Q parameter, defined as

\[Q_{Toomre}(r) = \frac{\bar{c}_{s}(r)\kappa(r)}{\pi \Sigma(r)},\]

where $\kappa$ is the epicyclic frequency and $\bar{c}_{s}$ is the RMS sound speed, is calculated using the above surface density, assuming a Keplerian rotation profile and a central star mass of unity (i.e., $\kappa(r) = \Omega(r)$, where $\Omega(r) = r^{-3/2}$). The sound speed for each particle $i$ is calculated from the stored thermal energy and $\gamma$ (ratio of specific heats) according to

\[\begin{split}c_{s,i}^{2} = \left\{ \begin{array}{ll} \frac23 u_{i}, & \gamma = 1; \\ (\gamma-1)\gamma u_i, & \gamma \neq 1; \end{array}\right.\end{split}\]

from which the RMS sound speed is calculated as the square root of the average of $c_{s}^{2}$ on the particles in the annulus.

Other useful functionality¶

Conversion of binary data files to ascii¶

splash has a command line feature which can be used to convert binary SPH dump files into ascii format. The syntax is

splash to ascii dump001 dump002 dump???

which will convert all of the dump files listed on the command line into ascii format (called dump001.ascii, dump002.ascii etc.), with columns as would be listed in the main menu if you opened the dump file in splash . Note that the output includes calculated extra quantities such as the radius if these have been turned on [in the d) menu] and the settings saved to the splash.defaults file. Similarly the data will be output in physical units if a splash.units file is present.

For other command line options, see Command line options.

Interpolation of SPH data to 2D and 3D grids¶

splash has a command line feature which can be used to read binary SPH dump files and output 3D gridded data in a variety of formats. The syntax is

splash to grid dump001 dump002 dump???

which will interpolate the density, velocity (if present) and magnetic field (if present) onto a 3D grid and output the results to files (the default output format is ascii, with one file for each quantity interpolated). Other data columns in the SPH file can be interpolated using the “allto” option, which interpolates all of the columns to the grid:

splash allto grid dump001 dump002 dump???

The grid interpolation uses the $x$, $y$, and $z$ limits — as saved to the splash.limits file — for the box, and the grid size is given by the “set number of pixels” option in the r)ender menu — as saved to the splash.defaults file. Automatic pixel determination also works (if npixels = 0) but there is a sensible upper limit placed on the grid size determined in this manner to avoid ridiculous memory/disk usage. Options can be set to change various aspects of the grid interpolation behaviour. For example you to use periodic boundary conditions use:

splash to grid dump001 --periodic

or alternatively you can set this as an environment variable:

export SPLASH_TO_GRID_PERIODIC=yes
splash to grid dump001

To change the number of pixels, you can use:

splash to grid dump001 --periodic --npix=100,100,000

and to interpolate only columns 4 and 5 you can use:

splash to grid dump001 --periodic --npix=100,100,000 --grid=1,5

For all possible output formats, use splash --help or see the full list of command line options in Command line options.

Using splash to calculate global quantities as a function of time.¶

splash has a command line feature that can be used to calculate global quantities on the particles as a function of time, for example kinetic, thermal, magnetic and total energy, total linear and angular momentum. An example to calculate the energies in a sequence of dump files is:

splash calc energies dump001 dump002 dump???

Other options are given by typing ’splash calc’, which currently has the following options:

splash calc energies     : calculate KE,PE,total energy vs time
                           output to file called 'energy.out'
       calc massaboverho : mass above a series of density thresholds vs time
                           output to file called 'massaboverho.out'
       calc max          : maximum of each column vs. time
                           output to file called 'maxvals.out'
       calc min          : minimum of each column vs. time
                           output to file called 'minvals.out'
       calc diff         : (max - min) of each column vs. time
                           output to file called 'diffvals.out'
       calc amp          : 0.5*(max - min) of each column vs. time
                           output to file called 'ampvals.out'
       calc delta        : 0.5*(max - min)/mean of each column vs. time
                           output to file called 'deltavals.out'
       calc mean         : mean of each column vs. time
                           output to file called 'meanvals.out'
       calc rms          : (mass weighted) root mean square of each column vs. time
                           output to file called 'rmsvals.out'
       calc timeaverage  : time average of *all* entries for every particle
                          output to file called 'time_average.out'
       calc ratio        : ratio of *all* entries in each file compared to first
                           output to file called 'ratio.out'

For the energies and massaboverho options to be successful, splash must be aware of the locations of the corresponding columns in the data (i.e., by the column identification given in the set_labels routine corresponding to the data read). For the massaboverho option an input file is required specifying the density thresholds (a default version is written if the appropriate file is not already present).

Using splash to time average a series of files¶

The splash calc timeaverage command line option (see Using splash to calculate global quantities as a function of time.) can be used to produce a time average of a series of files from any splash-readable format. This computes the time-average of every individual entry in the file as represented in splash as a table of rows (or ‘particles’) and columns (or ‘quantities defined on particles’). The output is an ascii file with the same rows and columns, averaged over all the snapshots on the command line. The number of columns is doubled in the output, giving the standard deviation for each quantity in the corresponding column (e.g., the standard deviation for column 1 is output in column $N + 1$).

Examples of how this could be use might be to produce the time-averaged power spectrum from a series of ascii files containing power spectra for individual output times, or the time averaged probability density function (PDF) from PDFs produced by splash (see sec:pdfs:).

The resulting ascii file, called time_average.out can be plotted using the ascii splash binary (asplash).

For other command line options, see Command line options.

Reading/processing data into images without having to answer prompts¶

Previously, the only way to run splash non-interactively was to write a small shell script which runs splash and answers the prompts appropriately. For example:

#!/usr/bin/tcsh
cd plot
splash myrun* << ENDINPUT
2
1
8
0
/png
q
ENDINPUT

which would plot the data in columns 2 and 1 and render the data in column 8 with output to file mypostscript.ps.

However, in more recent versions splash can be invoked with plot options on the command line. Thus to achieve the same as in the example given above we would simply use

splash myrun* -x 1 -y 2 -render 8 -dev /png

or simply

splash myrun* -r 8 -dev /png

which will assume sensible default values (2 and 1 respectively) for the y and x axes. Similarly a vector plot can be specified with -vec and a contour plot with -cont. The full list of command-line flags is given in Command line options.

If plotting options have been only partially specified on the command line, no prompts will appear for the remaining options and default values will be assumed for the remaining options. For example, the default device will be /xw giving an interactive plot.

Computing volume-weighted probability density functions from SPH data using SPLASH¶

The best way to compute a volume-weighted probability density function on SPH particles is to interpolate the density field to a grid and compute the histogram of the number of grid cells containing a given value of the desired quantity.

The grid2pdf utility included with splash can be used to compute the density PDF from gridded data output by the splash to grid utility (see Interpolation of SPH data to 2D and 3D grids).

To use this feature, you will need to output grids in “binary” format, e.g:

splash to gridbinary turb_00020

or if you want to skip the velocity interpolation (assuming density in column 6):

splash to gridbinary turb_00020 --grid=6

this produces a file called turb_00020.grid, then follow this with:

cd $SPLASH_DIR; make grid2pdf
cd -
$SPLASH_DIR/splash/bin/grid2pdf turb_00020.grid

which produces:

turb_00020.grid_pdf_ln_density.dat

this is just a two-column ascii file, so you can then plot this with your favourite plotting tool, e.g.:

splash -ev turb_00020.grid_pdf_ln_density.dat

Making frames across multiple processors¶

Making identical plots of a series of dump files for a movie is a task which can inherently be done in parallel. Included in the splash/scripts directory is a perl wrapper for splash (“splash_parallel.pl”) which distributes multiple instances of splash across multiple machines, either via ssh or using Apple’s xgrid, with a common input file as described in Reading/processing data into images without having to answer prompts. The limitation to this is that you need to have a disk which can be mounted from all client machines (i.e., they can read the data files) and preferably with password-less access (e.g. using an ssh key-exchange or Kerberos authentication). The script itself may need some slight adjustment for your particular system.

However, with large datasets often the slowest part of the rendering process can be reading the data file. A good way of crippling a system is therefore to set 100 jobs going which all decide to read a large data file from disk at the same time. To avoid this the script allows the user to set a delay between launching jobs (preferably slightly longer than the length of time it takes to read a single dump file), but some care is needed to avoid disaster. You have been warned!

Using special characters in the plot labels¶

Several of the examples shown in this manual use special characters (such as the $\int$ character) in the plot labels. In giza these can be specified using TeX-like escape sequences, or with the escape sequences used in pgplot. For example to plot the greek letter $\rho$ we would use

label = 'this would print the greek letter \rho'

or, in pgplot-style:

label = 'this would print the greek letter \gr'

where \gr is the pgplot escape sequence for $\rho$.

In giza , which uses real fonts rather than the bitmapped characters used in pgplot, special characters are implemented with unicode characters. Thus, you need to select a font that has the appropriate characters included. The font can be changed using the GIZA_FONT environment variable.

For other characters the procedure is similar. For example for the integral

\[\int v_x \mathrm{dx}\]

we would use the TeX-like expression

label = '\int v_x dx'

or equivalently, in pgplot-style

label = '\(2268) v\d x \u dx'

where \(2268) is the pgplot escape sequence for the integral sign. The \d indicates that what follows should be printed as subscript and \u correspondingly indicates a return to normal script (or from normal script to superscript). All of the escape sequences for special characters are listed in the appendix to the pgplot user guide.

WARNING: Note that the use of escape characters can be compiler dependent and may not therefore work on all compilers (for example the intel compiler needs the -nbs flag).

Outputting the raw pixel map to a file¶

The actual pixel map rendered to the graphics device (i.e., when a quantity is rendered to pixels, not for particle plots) can be output directly to a file, or series of files by using the -o command line option when you invoke splash. This is useful if you need to compare the image to the output from another code (e.g. using a different visualisation tool) or if you wish to have a “raw” rendering, that is without annotation on the plots.

Invoking splash with -o lists the currently implemented formats:

possible formats for -o option:
-o ppm   : dump pixel map to portable pixel map file
-o pfm   : dump pixel map to portable float map file
-o ascii : dump pixel map to ascii file

For example, to output the pixel map in ascii format, use:

splash discG_00300 -o ascii -r 6 -dev /png

giving:

> writing pixel map to file discG_00300_columndensitygcm2_proj.pix ...OK

This produces a file as follows:

$ more discG_00300_columndensitygcm2_proj.pix
# discG_00300_columndensitygcm2_proj.pix created by SPLASH
# Contains 2D pixel array 512 x 512 written as
#   do j=1,512
#      write(*,*) dat(1:512,j)
#   enddo
# column density [g/cm^2]: min =   9.697428E-12 max =   7.487661E+03
# x axis: min =  -4.000000E+03 max =   4.000000E+03
# y axis: min =  -4.000000E+03 max =   4.000000E+03
# 512 512
  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+00  0.000000E+0
...

The number of pixels in the image can be controlled using the set number of pixels’ option in the :ref:`sec:menu-r (making sure you save the settings to the splash.defaults file using the (s)aving your settings).

Reading raw pixel maps from splash into Python¶

See above for how to output the raw pixel map to a file. The resulting .pix file can be read into Python using the command:

array = np.loadtxt('discG_00300_columndensitygcm2_proj.pix',skiprows=9)
print (array.shape)

For other command line options, see Command line options.

Frequently Asked Questions¶

Here is a collection of the most common problems with using SPLASH. You may also wish to browse the archives of the splash-users discussion list. See also the old pgplot faq.

How do I make a movie from splash output?¶

At the graphics device prompt, choose /png. This will produce a series of files:

splash_0000.png
splash_0001.png
splash_0002.png
...

There are then many procedures for getting from these files to an animation. See Producing a sequence of plots for a movie

Using ffmpeg to make a movie from png files¶

With ffmpeg installed, you can use the script provided in splash/scripts/movie.sh, e.g.:

~/splash/scripts/movie.sh

This should produce a movie called movie.mp4 from your splash_*.png files. The contents of this script are just some options for ffmpeg that produce a nice movie:

#!/bin/bash
#
# script to make an mpeg4 movie from splash_*.png files
# requires ffmpeg utility
#
# DJP, Feb 2014
#
opts='-r 10 -vb 50M -bt 100M -vf setpts=4.*PTS'
ffmpeg -i splash_%04d.png $opts movie.mp4

The main options to play with are the bitrate (-bt 100M) that controls the quality and the -vf setpts=4.*PTS which controls how fast the movie plays (setpts=1.*PTS gives 10 frames per second, this is usually too fast so we slow it down).

The industry-standard compression codec is the H.264 MPEG codec which gives a nice small movie that is still good quality. The best way to reduce the movie size further is to restrict the data rate.

Important

The .mp4 files produced by ffmpeg as above currently not play directly in the Slack app. To ensure this you need to mandate that the pixel format is “yuv420p”, by adding the flag -pix_fmt yuv420p. You must also ensure that the page dimensions are given in even (not odd) numbers of pixels.

Can I make animated gifs?¶

Another option is to produce an animated gif using the convert utility that comes with the ImageMagick package. As an example (kindly provided by John Mansour), use:

convert -delay 10 -loop 0 splash_*.png animatedgif.gif

where delay gives a delay in ms between frames, and loop determines whether the animated gif will loop (0= infinite, 1=1 loop, etc). The problem with an animated GIF is that they are slow unless you have lots of memory. However it is easy to convert an animated GIF into other movie formats.

I’m getting a segmentation fault, is splash broken?¶

This could be a stacksize issue (particularly with large data reads). Try:

ulimit -s unlimited

before invoking splash.

How do I write a postscript/eps file for LaTeX?¶

See Producing figures for LaTeX documents. At the graphics device prompt, type:

Graphics device/type (? to see list, default /xw): /eps

choose /eps. This will write the output to a file called splash.eps in the current directory. Alternatively you can specify the filename explicitly by typing “myfile.eps”. The file can be incorporated directly into LaTeX documents.

How can I get a colour scheme like the one Joe Bloggs uses?¶

Send me an image (any sensible format will do) containing a colour bar using the scheme you want to rip off and I will add it into splash. Also feel free to play with my scripts for grabbing colour maps yourself.

What about boundaries? How does the rendering work near a boundary?¶

Usual practice in SPH simulations near boundaries is to introduce ghost particles which mirror the real particles. splash does not explicitly setup any ghost particles but will use any that are present in the data (see next question for how to specify multiple particle types). Additional particle types contribute to the rendering calculations but not to the determination of the plot limits. Note, however, that splash does not set up ghost particles itself, as this may depend on the type and location of the boundary. Thus if your simulation uses ghost particle boundaries, the ghost particles should be dumped alongside the gas particles in the output file so that their positions, masses, densities and smoothing lengths can be read into splash and used to render the image appropriately.

How does splash handle multiple particle types?¶

splash can handle up to 6 different particle types. These can be turned on and off in the particle plot (o)ptions. These types are be specified in the set_labels part of the read_data routine, which contains some lines of code along the lines of:

ntypes = 3
labeltype(1) = 'gas'
labeltype(2) = 'ghost'
labeltype(3) = 'sink'
UseTypeInRenderings(1) = .true.
UseTypeInRenderings(2) = .true.
UseTypeInRenderings(3) = .false.

which says that there are 3 particle types, with names as given, and that types 1 and 2 are SPH particles and should be used in the rendering where appropriate (i.e., only when plotting of this type is turned on in the particle plot (o)ptions). Particle types which are to be used in renderings should have masses, densities and smoothing lengths read. Non-SPH particle types (e.g. sink particles) can be optionally plotted on top of Rendered plots.

What does SPLASH stand for?¶

Urrmmm… it has SPH in it and it sounded good. I thought of:

“Some Pretty Little Application for Smoothed (particle) Hydrodynamics”
“Smoothed Particles Look Amazingly Stunning Here”
“So People Love Analysing Simulations of Hydrodynamics”
“Simulating Particles Like A Superfast Horse”

Your suggestions on a postcard please.

SPLASH is so great. Can I send you loads of money?¶

I accept donations in the form of citations to the SPLASH paper (Price, 2007, PASA, 24, 159-173). Just like sending cash, only… not. This may change if I am flooded with requests from people wanting to send large sums of money.

Contributing¶

Please contribute! All user contributions or suggestions are greatly appreciated. In particular, please send:

Bugs! (best way is to file an issue)
Feature requests/suggestions
Pretty pictures for the gallery.

If you are *really* keen, you may also like to consider:

Exact solution routines for test problem(s). Even just an analytic description from which I can write the code.
Suggestions/tips on possible visualisation techniques
More colour schemes (simply email me a table of the rgb colour indices, or failing that simply an image of the colour scheme and I will add it).

For discussion, feel free to raise an issue on the tracker.

Otherwise, you can send an email to the mailing list or get in touch directly.

Acknowledgements¶

Several of the routines were developed from ideas used by Matthew Bate and splash has been refined by many useful discussions with Matthew. The polytrope exact solution is from a routine by Joe Monaghan. I am indebted to one Thomas S. Ullrich at the University of Heidelberg who wrote the prompting module which is used throughout the program and to Roland Schmehl who wrote the excellent function parser module (made available at http://fparser.sourceforge.net). Last but not least, a huge thanks especially to all the users who have given feedback which has helped to improve splash including, but not limited to:

Stefan Adami, Craig Agnor, Richard Alexander, Gabe Altay, Pau Amaro-Seoane, Alessandro Atrani, Sumedh Anathpindika, Ben Ayliffe, Andreas Bauswein, Mark Bennett, Sergei Biriukov, David Brown, Florian Buerzle, Josh Calcino, Paul Cornwall, Jared Coughlin, Carlos Cuesta, Daniel Cunnama, Alan Duffy, Clare Dobbs, Carrie Elliot, Claude-Andr'e Faucher-Gigu`ere, Stefano Facchini, Juan Pablo Farias, Christoph Federrath, Laure Fouchet, Sergio Gelato, Thomas Grief, Doron Grossman, Jean-Franc{c}ois Gonzalez, Johnny Hitti, Mark Hutchison, Vid Irv{s}iv{c}, John Jones, Sky King, Laura Kreidberg, Guillaume Laibe, Ben Lewis, David Liptai, Giuseppe Lodato, Hayley Macpherson, David Madlener, John Mansour, Ruben Martin, Daniel Mentiplay, Farzana Meru, Andrew McLeod, Nick Moeckel, Shazrene Mohamed, Rebecca Nealon, Chris Nixon, Alex Pettitt, Cody Raskin, John Regan, Dave Rundle, Alison Sills, Kevin Sooley, Phil Sutton, Robert Thompson, St'even Toupin, Terry Tricco, Yusuke Tsukamoto, Sigfried Vanaverbeke, Enrique Vazquez-Semadeni Antonio Vazquez, Tim Waters, James Wurster, Matt Young.

And to everyone who has cited the splash paper!

References¶

DW94: Dai & Woodward (1994), J. Comp. Phys. 111, 354
MP04: Monaghan & Price (2004), MNRAS 350, 1449
Price07: Price (2007), PASA 24, 159
Price12: Price (2012), J. Comp. Phys. 231, 759
Price18: Price et al. (2018), PASA 35, e031
PB07: Price & Bate (2007), MNRAS 377, 77
PM04: Price & Monaghan (2004), MNRAS 348, 123
PM07: Price & Monaghan (2007), MNRAS 374, 1347
RP07: Rosswog & Price (2007), MNRAS 379, 915
RJ95: Ryu & Jones (1995), ApJ 442, 228
Toro92: Toro (1992), Phil. Trans. Phys. Sci. Eng. 338, 43

Links¶

Project homepage: http://users.monash.edu.au/~dprice/splash
Code repository: https://github.com/danieljprice/splash
Documentation: https://splash-viz.readthedocs.io
Code paper: https://adsabs.harvard.edu/abs/2007PASA…24..159P

Source code overview¶

Here is a brief and outdated description of various files making up the code:

Filename	Description

allocate.f90	allocates memory for main arrays
calc_quantities.f90	calculates additional quantities from particle data
colours.f90	colour schemes for rendering
colourparts.f90	colours particles
defaults.f90	writes/reads default options to/from file
exact.f90	module handling exact solution settings
exact_densityprofiles.f90	various $N-$body density profiles
exact_fromfile.f90	reads an exact solution tabulated in a file
exact_mhdshock.f90	some tabulated solutions for mhd shocks
exact_polytrope.f90	exact solution for a polytrope
exact_rhoh.f90	exact relation between density and smoothing length
exact_sedov.f90	exact solution for sedov blast wave
exact_shock.f90	exact solution for hydrodynamic shocks
exact_wave.f90	exact solution for a propagating sine wave
exact_toystar.f90	exact solution for the toy star problem
exact_toystar2D.f90	exact solution for the 2D toy star problem
get_data.f90	wrapper for main data read
geometry.f90	module handling different coordinate systems
globaldata.f90	various modules containing “global” variables
interactive.f90	drives interactive mode
interpolate1D.f90	interpolation of 1D SPH data to grid using kernel
interpolate2D.f90	interpolation of 2D SPH data to grid
interpolate3D_xsec.f90	3D cross section interpolations
interpolate3D_projection.f90	3D interpolation integrated through domain
legends.f90	plots (time) legend on plot
limits.f90	sets initial plot limits and writes to/reads from limits file
menu.f90	main menu
options_data.f90	sets options relating to current data
options_limits.f90	sets options relating to plot limits
options_page.f90	sets options relating to page setup
options_particleplots.f90	sets options relating to particle plots
options_powerspec.f90	sets options for power spectrum plotting
options_render.f90	sets options for render plots
options_vector.f90	sets options for vector plots
options_xsecrotate.f90	sets options for cross sections and rotation
particleplot.f90	subroutines for particle plotting
plotstep.f90	main “backbone” of the code which drives plotting of a single timestep
read_data_dansph.f90	reads data from my format of data files
read_data_mbate.f90	reads data from Matthew Bate’s format of data files
read_data_xxx.f90	reads data from …
render.f90	takes array of pixels and plots render map/contours etc
rotate.f90	subroutines controlling rotation of particles
setpage.f90	sets up the PGPLOT page (replaces call to PGENV/PGLAB)
splash.f90	main program, handles startup/ command line reading
timestepping.f90	controls stepping through timesteps
titles.f90	reads a list of titles to be used to label each timestep
transform.f90	applies various transformations to data (log10, 1/x, etc)

Coordinate transforms¶

Particle positions and vectors defined on the particles can be plotted in non-cartesian coordinate systems. The coordinate system can be set via the particle plot (o)ptions, via the change coordinate system option. The actual coordinate transformations are defined in a standalone Fortran module called geometry.f90 and the precise details can be determined by looking in this file. For reference, however the transformations are given below.

Cylindrical Polar Coordinates¶

For cylindrical coordinates the transformations are:

\[\begin{split}\begin{array}{lclp{1cm}lcl} R & = & \sqrt{x^2 + y^2} & & x & = & R\cos\phi \\ \phi & = & \tan^{-1}{(y/x)} &; & y & = & R\sin\phi \\ z & = & z & & z & = & z\\ \end{array}\end{split}\]

where vectors transform according to:

\[\begin{split}\begin{array}{lclp{1cm}lcl} v_R & = & v_x \frac{x}{R} + v_y \frac{y}{R} & & v_x & = & v_R \cos\phi - v_\phi \sin\phi \\ v_\phi & = & v_x \left(\frac{-y}{R}\right) + v_y \left(\frac{x}{R}\right) &; &v_y & = & v_R \sin\phi + v_\phi \cos\phi \\ v_z & = & v_z & & v_z & = & v_z. \\ \end{array}\end{split}\]

In the case where these vectors are velocities, the $v_{\phi}$ component corresponds to $v_{\phi} = r\dot{\phi}$. At the origin we assume $\phi = 0$, implying $\cos\phi = 1$ and therefore $\frac{x}{R} = 1$ and $\frac{y}{R} = 0$. This ensures the transformations are reversible everywhere.

Spherical Polar Coordinates¶

For spherical coordinates the transformations are:

\[\begin{split}\begin{array}{lclp{1cm}lcl} r & = & \sqrt{x^2 + y^2 + z^{2}} & & x & = & r\cos\phi\sin\theta\\ \phi & = & \tan^{-1}{(y/x)} &; & y & = & r\sin\phi\sin\theta \\ \theta & = & \cos^{-1}(z/r) & & z & = & r\cos\theta \\ \end{array}\end{split}\]

where vectors transform according to:

\[\begin{split}\begin{array}{lclp{1cm}lcl} v_r & = & v_x \frac{x}{r} + v_y \frac{y}{r} + v_{z}\frac{z}{r} & & v_x & = & v_r \cos\phi\sin\theta- v_\phi \sin\phi + v_\theta \cos\phi\cos\theta \\ v_\phi & = & v_x \left(\frac{-y}{\sqrt{x^2 + y^{2}}}\right) + v_y \left(\frac{x}{\sqrt{x^2 + y^{2}}}\right) &; &v_y & = & v_r \sin\phi\sin\theta + v_\phi \cos\phi + v_{\theta} \sin\phi\cos\theta \\ v_\theta & = & v_{x}\frac{xz}{r \sqrt{x^{2} + y^{2}}} + v_{y}\frac{yz}{r \sqrt{x^{2} + y^{2}}} - v_{z}\frac{(x^{2} + y^{2})}{r\sqrt{x^{2} + y^{2}}} & & v_z & = & v_r \cos\theta - v_\theta \sin\theta. \\ \end{array}\end{split}\]

In the case where these vectors are velocities, the components $v_{\phi}$ and $v_{\theta}$ correspond to $v_{\phi} = r\sin{\theta}\dot{\phi}$ and $v_{\theta} = r\dot{\theta}$ respectively.

Toroidal Coordinates¶

Toroidal coordinates represent a local frame of reference inside a torus. The coordinate transformations are given by

\[\begin{split}\begin{array}{lclp{1cm}lcl} r & = & \sqrt{[(x^2 + y^2)^{1/2} - R]^{2} + z^{2}} & & x & = & (r\cos\theta + R) \cos\phi \\ \theta & = & \tan^{-1} \left[\frac{z}{(\sqrt{x^{2} + y^{2}} - R)}\right] &; & y & = & (r\cos\theta + R)\sin\phi \\ \phi & = & \tan^{-1}(y/x) & & z & = & r\sin\theta \\ \end{array}\end{split}\]

where $R$ is the radius of the torus. The use of the inverse tangent in $\theta$ instead of $\theta = \sin^{-1}(z/r)$ is necessary to get the quadrant correct (via the atan2 function). Vectors transform according to:

\[\begin{split}\begin{array}{lclp{2cm}lcl} v_r & = & v_x \frac{x(r_{cyl} - R)}{r r_{cyl}} + v_y \frac{y(r_{cyl} - R)}{r r_{cyl}} + v_{z} \frac{z}{r} & & v_x & = & v_r \cos\theta\cos\phi- v_\theta \sin\theta\cos\phi - v_\phi\sin\phi \\ v_\theta & = & v_x \frac{-zx}{r r_{cyl}} + v_y\frac{-zy}{r r_{cyl}} + v_{z}\frac{(r_{cyl} - R)}{r} &; &v_y & = & v_r \cos\theta\sin\phi - v_\theta \sin\theta\sin\phi + v_\phi\cos\phi \\ v_\phi & = & v_{x} \left(\frac{-y}{r_{cyl}}\right) + v_{y} \left(\frac{x}{r_{cyl}}\right) & & v_z & = & v_{r}\sin\theta + v_{\theta} \cos\theta \\ \end{array}\end{split}\]

where we have defined, for convenience,

\[r_{\rm cyl} = \sqrt{x^{2} + y^{2}} = r\cos\theta + R. \nonumber\]

and the velocities $v_\theta$ and $v_\phi$ correspond to $r \dot{\theta}$ and $r_{\rm cyl} \dot{\phi}$, respectively. The torus radius $R$ is a parameter in the geometry module and is set to $1$ by default.

Flared Cylindrical Polar Coordinates¶

For flared cylindrical coordinates the transformations are:

\[\begin{split}\begin{array}{lclp{1cm}lcl} R & = & \sqrt{x^2 + y^2} & & x & = & R\cos\phi \\ \phi & = & \tan^{-1}{(y/x)} &; & y & = & R\sin\phi \\ z' & = & z \left(R_{\rm ref}/{R}\right)^\beta & & z & = & z' (R/R_{\rm ref})^\beta \\ \end{array}\end{split}\]

where $R_{\rm ref}$ is the reference radius and $\beta$ is the flaring index, both parameters that can be set by the user. Vectors transform according to:

\[\begin{split}\begin{array}{lclp{1cm}lcl} v_R & = & v_x \frac{x}{R} + v_y \frac{y}{R} & & v_x & = & v_R \cos\phi - v_\phi \sin\phi \\ v_\phi & = & v_x \left(\frac{-y}{R}\right) + v_y \left(\frac{x}{R}\right) &; &v_y & = & v_R \sin\phi + v_\phi \cos\phi \\ v_{z'} & = & -\beta \frac{xz}{R^2} \left(\frac{R_{\rm ref}}{R}\right)^\beta v_x -\beta \frac{yz}{R^2} \left(\frac{R_{\rm ref}}{R}\right)^\beta v_y + \left(\frac{R_{\rm ref}}{R}\right)^\beta v_z & & v_z & = & \beta \frac{z'}{R} \left(\frac{R}{R_{\rm ref}}\right)^\beta v_R + \left(\frac{R}{R_{\rm ref}}\right)^\beta v_{z'}. \\ \end{array}\end{split}\]

In the case where these vectors are velocities, the $v_{\phi}$ component corresponds to $v_{\phi} = r\dot{\phi}$.

Logarithmic Flared Cylindrical Polar Coordinates¶

For logarithmic flared cylindrical coordinates the transformations are:

\[\begin{split}\begin{array}{lclp{1cm}lcl} d & = & \log_{10} ( \sqrt{x^2 + y^2} ) & & x & = & R\cos\phi \\ \phi & = & \tan^{-1}{(y/x)} &; & y & = & R\sin\phi \\ z' & = & z \left(R_{\rm ref}/{R}\right)^\beta & & z & = & z' (R/R_{\rm ref})^\beta \\ \end{array}\end{split}\]

where $R_{\rm ref}$ is the reference radius and $\beta$ is the flaring index, both parameters that can be set by the user. Vectors transform according to:

\[\begin{split}\begin{array}{lclp{1cm}lcl} v_d & = & v_x \frac{x}{R}f^{-1} + v_y \frac{y}{R}f^{-1} & & v_x & = & f v_d \cos\phi - v_\phi \sin\phi \\ v_\phi & = & v_x \left(\frac{-y}{R}\right) + v_y \left(\frac{x}{R}\right) &; &v_y & = & f v_d \sin\phi + v_\phi \cos\phi \\ v_{z'} & = & -\beta \frac{xz}{R^2} \left(\frac{R_{\rm ref}}{R}\right)^\beta v_x -\beta \frac{yz}{R^2} \left(\frac{R_{\rm ref}}{R}\right)^\beta v_y + \left(\frac{R_{\rm ref}}{R}\right)^\beta v_z & & v_z & = & \beta \frac{z'}{R} \left(\frac{R}{R_{\rm ref}}\right)^\beta f v_d + \left(\frac{R}{R_{\rm ref}}\right)^\beta v_{z'}. \\ \end{array}\end{split}\]

where $R \equiv 10^d$ and correspondingly $d = \log_{10} R$ and $f \equiv R \ln (10)$.

Exact solutions¶

splash contains a library of exact or analytic solutions to common test problems that have been used extensively in benchmarking codes (e.g. [PM04], [RP07], [Price12], [Price18]).

Errors¶

The error norms calculated when exact solutions are plotted are as follows: The error for each particle is given by

\[e_i = f_i - f_{exact},\]

where the exact solution $f_{exact}(x)$ is the solution returned from the exact solution subroutines (with resolution adjustable in the exact solution options menu option) interpolated to the position of the current particle $x_i$ via a simple linear interpolation. The absolute $L_1$ error norm is simply the average of the errors across the domain, calculated according to

\[\Vert e \Vert_{L_1} = \frac{1}{N f_{max}} \sum_{i=1}^N \vert e_i \vert,\]

where $f_{max}$ is the maximum value of the exact solution in the region in which the particles lie (also only particles in the current plot are used) which is used to normalise the error estimate. A better error norm is the $L_2$ or Root Mean Square (RMS) norm given by

\[\Vert e \Vert_{L_2} = \left[\frac{1}{N} \left( \frac{1}{f_{max}^2} \sum_{i=1}^N \vert e_i \vert^2 \right)\right]^{1/2}.\]

Finally the maximum error, or $L_\infty$ norm is calculated according to

\[\Vert e \Vert_{L_\infty} = \frac{1}{f_{max}} {\rm max}_i \vert e_i \vert.\]

which is the most stringent error norm.

The inset plot of the individual particle errors shows the fractional deviation for each particle given by

\[e_{i,frac} = (f_i - f_{exact}) / f_{exact}.\]

Shock tubes (Riemann problem)¶

The subroutine exact_shock plots the exact solution for a one-dimensional shock tube (Riemann problem). The difficult bit of the problem is to determine the jump in pressure and velocity across the shock front given the initial left and right states. This is performed in a separate subroutine (riemannsolver) as there are many different methods by which this can be done (see e.g. [Toro92]). The actual subroutine exact_shock reconstructs the shock profile (consisting of a rarefaction fan, contact discontinuity and shock, summarised in Fig. 29), given the post-shock values of pressure and velocity.

Example of exact solution for one-dimensional shock tube problem (red line) compared to the SPH solution (black line/particles), utilising the exact solutions incorporated in splash¶

The speed at which the shock travels into the ‘right’ fluid can be computed from the post shock velocity using the relation

\[v_{shock} = v_{post}\frac{(\rho_{post}/\rho_R)}{(\rho_{post}/\rho_R)- 1},\]

where the jump conditions imply

\[\frac{\rho_{post}}{\rho_R} = \frac{(P_{post}/P_R) + \beta}{1 + \beta (P_{post}/P_R)}\]

with

\[\beta = \frac{\gamma - 1}{\gamma + 1}.\]

Riemann solver¶

The algorithm for determining the post-shock velocity and pressure is taken from [Toro92].

Polytrope¶

The subroutine exact_polytrope computes the exact solution for a static polytrope with arbitrary $\gamma$. From Poisson’s equation

\[\nabla^2 \phi = 4\pi G \rho,\]

assuming only radial dependence this is given by

()¶\[\frac{1}{r^{2}} \frac{d}{dr} \left(r^{2} \frac{d\phi}{dr} \right) = 4\pi G \rho(r).\]

The momentum equation assuming an equilibrium state (${\bf v} = 0$) and a polytropic equation of state $P = K\rho^{\gamma}$ gives

()¶\[\frac{d\phi}{dr} = - \frac{\gamma K}{\gamma-1}\frac{d}{dr} \left[\rho^{(\gamma -1)} \right]\]

Combining (Eq. 1) and (Eq. 2) we obtain an equation for the density profile

()¶\[\frac{\gamma K}{4\pi G (\gamma - 1)} \frac{1}{r^{2}} \frac{d}{dr} \left[r^{2} \frac{d}{dr}\left( \rho^{\gamma-1} \right) \right] + \rho(r) = 0.\]

This equation can be rearranged to give

\[\frac{\gamma K}{4\pi G (\gamma - 1)} \frac{d^2}{dr^2} \left[r\rho^{\gamma-1}\right] + r\rho = 0.\]

The program solves this equation numerically by defining a variable

\[\mathcal{E} = r \rho^{\gamma-1}\]

and finite differencing the equation according to

\[\frac{\mathcal{E}^{i+1} - \mathcal{E}^i + \mathcal{E}^{i-1}}{(\Delta r)^2} = \frac{4\pi G (\gamma - 1)}{\gamma K} r \left(\frac{\mathcal{E}}{r}\right)^{1/(\gamma-1)}.\]

Linear wave¶

The subroutine exact_wave simply plots a sine function on a given graph. The function is of the form

\[y = \sin{(k x - \omega t)}\]

where $k$ is the wavenumber and $\omega$ is the angular frequency. These parameters are set via the input values of wavelength $\lambda = 2\pi/k$ and wave period $P = 2\pi/\omega$.

Input parameters for the linear wave exact solution¶
$\lambda$	wavelength
$P$	period

Sedov blast wave¶

The subroutine exact_sedov computes the self-similar Sedov solution for a blast wave.

Toy stars¶

The subroutine exact_toystar1D computes the exact solutions for the ‘Toy Stars’ described in [MP04]. The system is one dimensional with velocity $v$, density $\rho$, and pressure $P$. The acceleration equation is

\[\frac{dv}{dt} = - \frac{1}{\rho} \frac{\partial P}{\partial x} - \Omega^2 x,\]

We assume the equation of state is

\[P = K \rho^\gamma,\]

The exact solutions provided assume the equations are scaled such that $\Omega^2 = 1$.

Static structure¶

The static structure is given by

\[\bar \rho = 1- x^2,\]

Linear solutions¶

The linear solution for the velocity is given by

\[v = 0.05 C_s G_n(x) \cos{\omega t} ).\]

The density is

\[\rho = \bar{\rho} + \eta,\]

where

\[\eta = 0.1 C_s \omega P_{n+1}(x) \sin{(\omega t)}.\]

Non-linear solution¶

In this case the velocity is given by

\[v = A(t) x,\]

while the density solution is

\[\rho^{\gamma -1} = H(t) - C(t) x^2.\]

where the parameters A, H and C are determined by solving the ordinary differential equations

\[\begin{split}\begin{aligned} \dot{H} & = & -AH(\gamma -1), \\ \dot{A} & = & \frac{2K \gamma}{\gamma -1} C - 1 - A^2 \\ \dot{C} & = & -AC(1+ \gamma),\end{aligned}\end{split}\]

The relation

()¶\[A^2 = -1 - \frac{2 \sigma C}{\gamma -1} + kC^{\frac{2}{\gamma +1}},\]

is used to check the quality of the solution of the differential equations by evaluating the constant $k$ (which should remain close to its initial value).

MHD shock tubes¶

These are some tabulated solutions for specific MHD shock tube problems at a given time taken from the tables given in [DW94] and [RJ95].

h vs $\rho$¶

The subroutine exact_hrho simply plots the relation between smoothing length and density, i.e.,

\[h = h_{\rm fact} \left(\frac{m}{\rho}\right)^{1/\nu}\]

where $\nu$ is the number of spatial dimensions. The parameter $h_{\rm fact}$ is output by the code into the header of each timestep. For particles of different masses, a different curve is plotted for each different mass value.

Data reads and command line options¶

Requesting a customised data reader¶

The splash binary will read many formats. The default format is ascii or csv any ascii or csv data file where columns correspond to different quantities and rows correspond to each particle (actually I use splash to plot graphs for nearly all data in this form, whether SPH or not) – it will also sensibly skip header lines which do not have the same number of columns.

However, it is ultimately desirable to use SPLASH to directly visualise the (binary) output of your code. If your format is not amongst those distributed:

splash -f

then BEFORE you start writing your own routine, please consider whether or not a routine to read your format would be of more general use (e.g. to other users of your code). If so, PLEASE email me to request a new read_data routine for your format, by sending an email attaching:

an example dump

and

the source code from the routine which wrote the dump file.

Then I can write a read for your format that can be added to the SPLASH repository and distributed in all future versions. Whilst I aim never to change the interface to the read_data routines, it is not impossible that some changes may occur somewhere down the line (or enhanced functionality – for example the more advanced data reads are able to read only the required columns for a given plot from the file, rather than the whole file). It doesn’t matter if your code only has one user, we are still happy to do this as it makes splash more widely useable and saves trouble later.

Writing your own data read¶

If you really want to hack one yourself it is best to look at some of the other examples and change the necessary parts to suit your data files. Note that reading directly from unformatted data files is much faster than reading from formatted (ascii) data.

If you do end up writing your own, again, please email me the end result so I can add it to the officially supported data reads. This also makes it much easier for you to upgrade to newer versions as you do not require a locally customised version.

The second best way is to attempt to modify one of the existing data reads. Even then, there are some things to note: Most important is that, for the rendering routines to work, the density, particle masses and smoothing lengths for all of the (gas) particles must be read in from the data file and their locations in the main data array labelled using the integer parameters irho, ipmass and ih. Labelling of the location of other particle quantities (e.g. iutherm for the thermal energy) is used in order to plot the exact solutions on the appropriate graphs and also for calculating additional quantities (e.g. calculation of the pressure uses iutherm and irho).

The positions of vector components in the data columns are indicated by setting the variable iamvec of that column equal to the first component of the vector of which this component is a part. So if column 4 is a vector quantity (say ${\bf v}$ in 3D), then iamvec(4) = 4, iamvec(5) = 4 and iamvec(6) = 4. Similarly the string labelvec should be set, i.e., labelvec = 'v' for these columns.

Input parameters for the linear wave exact solution¶
\(\lambda\)	wavelength
\(P\)	period

Visualisation of smoothed particle hydrodynamics data using splash¶

Introduction¶

What it does¶

What it doesn’t do¶

splash, the paper¶

Version History¶

License¶

Getting started¶

Installing from your package manager¶

Stable version¶

Development version¶

Compiling from source¶

Basic instructions¶

Installing in your home space¶

Advanced installation guide¶

Fortran compilers¶

Cairo graphics library¶

Compiling and linking with giza¶

Reading your data¶

Command line options¶

Options affecting all data reads¶

Ascii data read¶

GADGET data read¶

VINE data read¶

sphNG data read¶

dragon data read¶

Stephan Rosswog data read¶

ndspmhd data read¶

H5Part data read¶

Environment variables¶

Changing the font¶

Endian changing¶

Basic splash usage¶

Simple two column plot¶

Interactive mode¶

Rendered plots¶

Cross section¶

Vector plots¶

Contour plots¶

Moving forwards and backwards through data files¶

Zooming in and out / changing plot limits¶

Producing figures for LaTeX documents¶

Producing a sequence of plots for a movie¶

Producing a movie from a sequence of images¶

Hollywood mode¶

Remote visualisation¶

Ten quick hints for producing good-looking plots¶

Changing plot settings¶

Menus¶

(s)aving your settings¶

set (m)ultiplot¶

Plotting more than one column from the same file on the same page (multiplot)¶

Plotting each particle type in a different panel (multiplot)¶

(d)ata options¶

Re-reading the initial data / changing the dump file¶

Using only a subset of data files / plotting every \(n-\)th dump file¶

Plotting more than one file without re-reading the data from disk¶

Calculating additional quantities not dumped¶

Plotting data in physical units¶

Rescaling data columns¶

Changing the default column labels¶

Plotting column density in g/cm\(^{2}\) without having x,y,z in cm¶

Changing physical unit settings¶

Changing the axis label to something like \(x\) \([ \times 10^{4} ]\)¶

Changing the time units¶

(i)nteractive mode¶

Adapting the plot limits¶

Making the axes logarithmic¶

Cycling through data columns interactively¶

Colouring a subset of the particles and retaining this colour through other timesteps¶

Working out which particles formed a particular object in a simulation¶

Plotting only a subset of the particles¶

Rendering using only a subset of the particles¶

Tracking a set of particles through multiple timesteps¶

Taking an oblique cross section interactively¶

(p)age options¶

Overlaying timesteps/multiple dump files on top of each other¶

Plotting results from multiple files in the same panel¶

Plotting more than one dump file on the same page¶

Changing axes settings¶