[WIP] Basic implementation of the formal integral #726
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Looks like a good starting point! Can you make a list of milestones to define the scope of this PR? |
|
Something like this? |
|
Well, if you want to do everything in one go then I'd say yes. |
|
Not necessarily. We can split the PR in several smaller and more focused ones. Do you have suggestions? |
|
I'd rather do it in smaller chunks and split it up into sensible sub-projects. E.g.:
|
|
According to your suggestion part 1 is mostly done? Only function documentation is missing as of right now. Additionally I'm not happy with the implementation as it does not allow for modularity right now. But I have changes in mind that should solve this. |
yeganer
added a commit
to yeganer/tardis
that referenced
this pull request
Jun 3, 2017
The refactor changes the way a spectrum is handled. Instead of creating a spectrum and updating it after each montecarlo iteration, we create a new spectrum from the MontecarloRunner whenever it is needed. Benefits: - TARDISSpectrum is now a general purpose spectrum class that is not linked to the Iteration sheme anymore. - It is now possible to save spectra for all iterations without them being updated each iteration. - This class can now also be used for spectra created by the formal integral approach (upcoming PR tardis-sn#726)
This was referenced Jun 3, 2017
Merged
|
|
yeganer
added a commit
to yeganer/tardis
that referenced
this pull request
Jun 7, 2017
The refactor changes the way a spectrum is handled. Instead of creating a spectrum and updating it after each montecarlo iteration, we create a new spectrum from the MontecarloRunner whenever it is needed. Benefits: - TARDISSpectrum is now a general purpose spectrum class that is not linked to the Iteration sheme anymore. - It is now possible to save spectra for all iterations without them being updated each iteration. - This class can now also be used for spectra created by the formal integral approach (upcoming PR tardis-sn#726)
This implementation is siginficantly faster than the python version.
This reduces the time to traverse the line list by a huge amount for a huge number of lines, especially if the start is near the red end of the line list.
Currently the c function expects a storage model, a blackbody temperature, an array of frequencies, the length of said array, the source function and the number of integration points. The function returns a array with the same length as the frequency array. Remember to free the result to prevent memory leaks. The python wrapper expects a simulation object, a np.ndarray of frequencies and the number of integration points. Additionally the identation of the C code was changed
Expand documentation Fix bug in populate_z.
In very rare cases when the outermost p line has an equal distance to the center as the radius of the outer shell, the code could try to access memory out of bounds.
yeganer
force-pushed
the
formal_integral
branch
2 times, most recently
from
c5be218 to
532d85d
Compare
The class stores a reference to model, plasma and runner to be able to calculate the formal integral. In this implementation it is first instantiated in MontecarloRunner.run() and available thereafter. It returns a TARDISSpectrum of the integrated spectrum.
The FormalIntegrator is responsible for all calculations associated with the formal integration method. It contains methods to calculate the source function and a wrapper to the c implementation.
There is a new option under spectrum which defaults to 'virtual' but can be set to 'real' or 'integrated' as well.
|
Close in favor of #740. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is still WIP and contains a basic implementation of the formal integral approach ported to C.
Milestones: