Compute and save the ISRF in every cell#231

Merged

astrofrog merged 40 commits into

hyperion-rt:mainfrom

dnarayanan:master

Jul 21, 2026

Commits on Jul 21, 2026

strawperson architecture in for computing specific energy as a function of frequency. note, this has a manual setting of frequencies and right now only computes, not saves

dnarayanan
authored and
astrofrog
committed
specific_energy_sum_nu is in all places specific_energy_sum is now i think

dnarayanan
authored and
astrofrog
committed
code now saves a ISRF in every cell. notably: AMR is still broken, and m.read is broken still with the new specific_energy_sum_nu array dimensions

dnarayanan
authored and
astrofrog
committed
specific_energy_sum_nu saves and reads. AMR still needs to be dealt with

dnarayanan
authored and
astrofrog
committed
grid_io_amr_template updated but commetned out since i cant quite get it to compile

dnarayanan
authored and
astrofrog
committed
grid modules now takes the n_isrf_wavelengths from the dust files

dnarayanan
authored and
astrofrog
committed
outputting the ISRF frequency bins

dnarayanan
authored and
astrofrog
committed
can now call whether ISRF is computed and saved via m.compute_isrf(True)

dnarayanan
authored and
astrofrog
committed
getting rid of internal comments showing where ive added new lines of code

dnarayanan
authored and
astrofrog
committed
some bugs corrected fixing the ISRF calculation. but some issues remain now when running without ISRF calculation turned on

dnarayanan
authored and
astrofrog
committed
bugs fixed that kept code from running when ISRF is turned off

dnarayanan
authored and
astrofrog
committed
getting rid of read_grid_5d since its not actually used

dnarayanan
authored and
astrofrog
committed
one more 5d file turned off

dnarayanan
authored and
astrofrog
committed
critical bug fix changing indexes starting from 0 to 1 [which caused, naturally, random segfaults]

dnarayanan
authored and
astrofrog
committed
Always accumulate specific_energy_sum so non-ISRF runs are heated correctly
astrofrog
committed
Guard ISRF accumulation behind compute_isrf and a dust check so dustless models no longer segfault
astrofrog
committed
Write ISRF frequency bins as a 1D dataset instead of a grid array to avoid an out-of-bounds segfault
astrofrog
committed
Fix typo in ISRF config reader method name so models can be read back
astrofrog
committed
Cache the ISRF frequency grid at setup and stop accumulating it during the final imaging iteration
astrofrog
committed
Remove a leftover comment and extra blank lines from the ISRF work
astrofrog
committed
Only populate the cached ISRF frequency grid when ISRF is enabled so dustless models do not segfault
astrofrog
committed
Fix the AMR write_grid_5d template slice so a regenerated grid_io_amr matches the working code
astrofrog
committed
Add regression tests for the ISRF covering heating, energy conservation, frequency bins, dustless models, AMR, and MPI
astrofrog
committed
Document the compute_isrf option and the specific_energy_nu and ISRF_frequency_bins outputs
astrofrog
committed
Allow the ISRF frequency grid to be set independently of the dust grid and bin photons in log space
astrofrog
committed
Add a frequencies argument to compute_isrf for a custom ISRF frequency grid
astrofrog
committed
Test and document the configurable ISRF frequency grid
astrofrog
committed
Expose specific_energy_nu through get_quantities and skip the ISRF frequency metadata in grid quantity reads
astrofrog
committed
Test and document retrieving specific_energy_nu via get_quantities
astrofrog
committed
Add a Voronoi ISRF regression test
astrofrog
committed
Skip the frequency-resolved ISRF quantity when exporting a grid to yt
astrofrog
committed
Control the frequency-resolved specific energy via output_specific_energy_nu and drop the ISRF naming
astrofrog
committed
Update config IO tests for output_specific_energy_nu and the frequency-grid setter
astrofrog
committed
Drop a redundant phrase from the frequency-resolved specific energy docs
astrofrog
committed
Clarify that specific_energy_nu frequencies are bin centers not edges
astrofrog
committed
Rename specific_energy_nu to specific_energy_spectrum throughout
astrofrog
committed
Warn in the docs that the spectrum end bins collect all out-of-range flux
astrofrog
committed
Skip the MPI spectrum test when the MPI build of the binaries is unavailable
astrofrog
committed
grid_integrate: deposit the specific energy spectrum at interaction points

authored and
astrofrog
committed
with specfic_type = 'additional', the frequency resolved specific energy array was copied into specific_energy_Additional_nu before ever being written...sometimes the unitialized memory was re-adde…

dnarayanan
authored and
astrofrog
committed