xpsi-group / xpsi Goto Github PK

X-PSI: X-ray Pulse Simulation and Inference

License: Other

Python 54.01% C 0.02% Cython 43.95% TeX 2.02%

astrophysics astronomy astronomical-algorithms astrophysical-simulation sampling parameter-estimation statistics-modeling modeling likelihood-functions posterior-sampling

xpsi's Introduction

X-PSI

An open-source package for neutron star X-ray Pulse Simulation and Inference.

X-PSI is designed to simulate rotationally-modified (pulsed) surface X-ray emission from neutron stars, taking into account relativistic effects on the emitted radiation. This can then be used to perform Bayesian statistical inference on real or simulated astronomical data sets. Model parameters of interest may include neutron star mass and radius (useful to constrain the properties of ultradense nuclear matter) or the system geometry and properties of the hot emitting surface-regions. To achieve this, X-PSI couples code for likelihood functionality (simulation) with existing open-source software for posterior sampling (inference).

It provides the following functionality:

Simulation of time- and energy resolved X-ray emission (pulse profiles) from the surfaces of neutron stars.
The facility to implement multiple models for surface patterns, atmospheres, and different instruments.
Coupling of pulse simulation functionality to statistical sampling software to infer spacetime and geometric parameters from pulse profile data.
An extensive suite of post-processing software to visualize posteriors and measures of model quality.

For more details on current and planned capabilities, check out the X-PSI documentation.

Installation and Testing

X-PSI has a complex set of dependencies, and is therefore currently best installed from source. The documentation provides step-by-step installation instructions for Linux and for limited MacOS systems.

Documentation

The documentation for XPSI, including a wide range of tutorials and scripts for running XPSI on HPC systems, can be found at https://xpsi-group.github.io/xpsi/.

How to get in touch or get involved

We always welcome contributions and feedback! We are especially interested in hearing from you if * something breaks, * you spot bugs, * if there is missing functionality, or you have suggestions for future development,

To get in touch, please open an issue. Even better, if you have code you'd be interested in contributing, please send a pull request (or get in touch and we'll help guide you through the process!).

For more information, you can take a look at the documentation's Contributing page. Please also make sure you take a look at the Code of Conduct.

Citing XPSI

If you find this package useful in your research, please provide the appropriate acknowledgment and citation. Our documentation provides more detail, including links to appropriate papers and BibTeX entries.

Copyright and Licensing

Legacy

An earlier version (pre-v0.5) of this project was named: A prototype open-source package for neutron star X-ray Pulsation Simulation and Inference.

xpsi's People

Contributors

Stargazers

Watchers

xpsi's Issues

Indicate which parameter is missing

xpsi/xpsi/ParameterSubspace.py

Line 134 in 6b7c8d8

raise KeyError('No parameter in subspace with matching name.')

xpsi/xpsi/ParameterSubspace.py

Line 221 in 6b7c8d8

raise KeyError('No parameter in subspace with matching name.')

When these KeyErrors show up, they do not indicate which is the parameter that isn't present in the subspace with a matching name.

Landing page for tutorials

The idea of this landing page is to guide new users in which tutorials they should do depending on their goals. I would also add additional pointers for other useful parts of the documentation.

Add docstrings and explainer comments to `hot.pyx`, `numerical.pyx`

These cython codes are complicated and uncommented. Update for better understandability of code

Add installation instructions for Helios and example job scripts.

Add installation instructions for Helios.

change mention of master to main in documentation

We have changed the default branch to main. this need to be updated everywhere.

Fix any broken tutorials

Also fix typos and overly complicated sentences.

global argument missing add_cede_temperature_prior method of module_generator.py

global argument missing for module variable in line 2480 within add_cede_temperature_prior method of module_generator:

xpsi/xpsi/module_generator.py

Line 2480 in 4ef9c1d

module += (

code in examples folder is broken

Hi all, I've been poking around XPSI, and especially I'd like to synthesize some data based on the model. I've found some code to do this in the examples folder, but I suspect it might be outdated? It refers to functions and classes that don't seem to exist in XPSI. Specifically, I tried running synthesise.ipynb, but it refers to a synethesise function that I can't find in the module where it's supposed to be (there are two separate functions there, and I don't know which one to use), and it also refers to a xpsi.Pulse module that I also can't find.

At this point I stopped trying to make it run. I think this would be worth checking/updating/adding to the official documentation?

Second component of the dual temperature region added twice if caching turned on?

It seems that the ceding region of a dual temperature region is added twice if caching is turned on for the signal object. The problem is in Signal.py the following line (295):

xpsi/xpsi/Signal.py

Line 295 in c904dce

signal = component

Here signal = component should be replaced by signal = component.copy() in order not to modify the original hot region components, when summing over them is done just for the caching purpose. Otherwise, the second component of the dual temperature region will be added to the signal again, when looping over the integrated signal.

UltraNest support

Would you be interested to add support for https://johannesbuchner.github.io/UltraNest/ ?
The interface should be very similar to pymultinest.

UltraNest is a very reliable tuning-parameter-free algorithm. I have published some examples where the UltraNest algorithm is unbiased while pymultinest gives a different answer. UltraNest is a pure-python package and very easy to install with pip or conda.

UltraNest also supports resuming from disk and MPI parallelisation, if that is useful to you.

Create script to switch between Numerical and Blackbody

copy files (hot', 'elsewhere') from surface_radiation_field/archive/' into 'surface_radiation_field' and recompile.

emcee compatibility

emcee has been upgraded to version 3.1.1, which requires the package autocorr, which is not compatible with python2.7.
So it is recommended to install emcee with pip install emcee==3.0.1.
A note about this could be added to the installation tutorial.

`b_f` which determines the size extent of the `omit` region for CST or `cede` region for CDT is set to twice the size of the `super` region

xpsi/xpsi/module_generator.py

Line 2566 in bd8ccc4

self.b_f = 2.0

If b_f is allowed to be greater than 1 for CST, then the omit region will outgrow the super region and will lead to no radiation being emanated from this region.
If b_f is allowed to be greater than 1 for CDT, then the cede region will outgrow the super region, and essentially lead to all emission coming from the cede and thus behaving as a ST region in these situations.

Set up Continuous Integration

Set up CI for X-PSI so that we can run tests automatically.

Potential cryptic error in Signal.py when data.channels and instrument.channels[a:b] have different size

In Signal.py (lines 95-99), there is a check that the data.channels are identical to the instrument.channel (with the data.index_range):

xpsi/xpsi/Signal.py

Line 96 in c904dce

if (data.channels != instrument.channels[a:b]).any():

However, in the case where the user does not define the instrument response and the data in the same channel range, the comparison data.channels != instrument.channels[a:b] will return a single False boolean. This also shows a warning (only once in python notebooks), that elementwise comparison will be deprecated. But in python2.7, this will remain the case that an element wise comparison of arrays of different size will return False and this raises a problem with the any() function since boolean.any() is not permitted. The error is:

AttributeError: 'bool' object has no attribute 'any'

So a problem of channels definition turns into a cryptic error for the user. Therefore, I suggest adding an extra check on the size of data.channels and instrument.channel with the data.index_range, before the comparison between the arrays is performed. Something simple, like:

a, b = data.index_range
if (len(data.channels) != len(instrument.channels[a:b]):
     raise ChannelError('Size of the channel array declared for event data does not match the size of the channel array declared for the loaded instrument response (sub)matrix.')

`photosphere.integrate` complains about missing `_spacetime` object if none of the `HotRegion` object parameters are left free.

Error message:

.../xpsi/Photosphere.pyc in integrate(self, energies, threads)
    322 
    323             if self._hot is not None:
--> 324                 self._signal = self._hot.integrate(self._spacetime,
    325                                                    energies,
    326                                                    threads,

AttributeError: 'CustomPhotosphere' object has no attribute '_spacetime'

When Star is instantiated, the following line should be called via __init__:

xpsi/xpsi/Star.py

Line 46 in b0ab3c6

photosphere.spacetime = self._spacetime

which should ensure that Photosphere also has the spacetime object, contrary to the error.

Calling star.update prior to calling photosphere.integrate has no effect on the error. In fact, oddly star.update doesn't yield any error in itself when executed even though it's supposed to in-turn call photosphere.embed which also requires photosphere._spacetime to be present. However, calling photosphere.embed manually does lead to the same error.
This could be a consequence of the if clause present (link below) before photosphere.embed when executing star.update.

xpsi/xpsi/Star.py

Line 94 in b0ab3c6

if photosphere.needs_update or self._spacetime.needs_update:

This in-turn suggests missing needs_update argument either in Photosphere or Star.

photosphere.embed subsequently calls hot.embed. Manually calling hot.embed to avoid missing photosphere._spacetime yields an error mentioning that fast_mode is not defined in the following line:

xpsi/xpsi/HotRegion.py

Line 949 in b0ab3c6

if self.fast_mode and not self.do_fast:

Manually passing spacetime object to photosphere also ends up complaining about the missing fast_mode.

Lower background limit set to have non-zero value if there are 0 background counts in a given channel

xpsi/xpsi/module_generator.py

Line 1552 in 71c310f

support[i,0] = support[j,1]

The upper limit support[i,1] is to be changed to a value > 0 when the background drops to 0, and not the lower limit.

Patch the background signal handling

Here the custom background implementation should store the background using the incident background setter:

https://github.com/ThomasEdwardRiley/xpsi/blob/2ce34f30abeff3dd1b3abadd7644f0a274f4fff4/xpsi/Background.py#L36

Then in the Signal class, the incident background property should be accessed.

Compute credible intervals for multiple runs supplied to `PostProcessor`

Right now, when using the CornerPlotter, the credible interval is plotted only for the first run in the IDs provided. The objective is to find a way to at least compute the other ones and save them as a dictionary with the run_ID as the key.

Python3 conversion

X-PSI should really support Python3 in the near future.

A work-in-progress PR exists in #49 and is ready for a first round of testing.

Citations for dependencies

Dear all,

As you know, developing software is a underappreciated activity in astronomy research. The only reward we can obtain are citations. In the documentation, could you please encourage users to also cite the packages xpsi is built upon?
Here is an example I have used: https://johannesbuchner.github.io/BXA/#citing-bxa-correctly
pymultinest and multinest seem to be relevant in xpsi, and its users, for example https://arxiv.org/pdf/2104.06928.pdf .

Cheers,
Johannes

Add install instructions for different systems

For example: on supercomputer CALMIP

Attribute error raised

https://github.com/ThomasEdwardRiley/xpsi/blob/753527d46fcd6d19186d2be468d6ecc0f2304a56/xpsi/Instrument.py#L94

Error message below:

Traceback (most recent call last):
  File "main_Tom_comparison.py", line 102, in <module>
    channel_edges=args.channels_path)
  File "/home/devarshi/anaconda3/envs/xpsi/lib/python2.7/site-packages/xpsi-0.5.3-py2.7-linux-x86_64.egg/xpsi/__init__.py", line 59, in decorator
    final = func(*args, **kwargs)
  File "/home/devarshi/msc-thesis/J0437_updated_instr_data_prior/STU/CustomInstrument.py", line 101, in from_SWG
    return cls(channel_edges[30:300, 1], RSP, edges, channels, beta)
  File "/home/devarshi/msc-thesis/J0437_updated_instr_data_prior/STU/CustomInstrument.py", line 19, in __init__
    super(CustomInstrument, self).__init__(*args)
  File "/home/devarshi/anaconda3/envs/xpsi/lib/python2.7/site-packages/xpsi-0.5.3-py2.7-linux-x86_64.egg/xpsi/Instrument.py", line 94, in __init__
    self.channel_edges = channel_edges
AttributeError: can't set attribute

channel_edges takes in rmf_nicer_channel_energies.txt

Add the simple test run scripts with synthetic data to examples directory

Provide updated example scripts and remove(?) the outdated ones.

[Feature] Elsewhere.py accepting dictionary

For consistency with other classes of the star (e.g. HotRegion), I think that Elsewhere.py should accept a dictionary even if it has only one parameter elsewhere_temperature.

That would mean changing:

            T = Parameter('elsewhere_temperature',
                          strict_bounds = (3.0, 7.0), # very cold --> very hot
                          bounds = bounds,
                          doc = 'log10 of the effective temperature elsewhere',
                          symbol = r'$\log_{10}(T_{\rm EW}\;[\rm{K}])$',
                          value = value)

into

            T = Parameter('elsewhere_temperature',
                          strict_bounds = (3.0, 7.0), # very cold --> very hot
                          bounds = bounds.get('elsewhere_temperature', None),
                          doc = 'log10 of the effective temperature elsewhere',
                          symbol = r'$\log_{10}(T_{\rm EW}\;[\rm{K}])$',
                          value = values.get('elsewhere_temperature', None))

change primary branch from 'master' to 'main

To be in-line with current standards across GitHub and elsewhere.

Sporadic HotRegion.RayError for certain parameter vectors

Calculating the likelihood for the same parameter vector gives some times HotRegion.RayError and sometimes not. For example, for a PSTU model with these parameters:

ray_error_vectors = np.array([[1.3728983226417986, 9.582564795756147, -0.7370097897513533, 0.020341921535536933, 0.2446314271154956, 0.3085074478486652, 6.161726551051954, 0.2722385280880412, 0.1088424612562043, 0.09511118603417756, -0.20397333010845675, 0.2730089443679007, 0.07205267520343456, 6.268835877470614, 0.13892737773308714, 0.13929302462200066, -0.6673219825864303, 43.29572745711499, 0.15926106535697776],
                             [1.3948493394100352, 9.885809466634319, -0.7378059269623075, 0.05209046240242854, 0.26593829571027794, 0.30030074135308593, 6.152591362014215, 0.3028962567969266, 0.11189861364104065, 0.3163235156902475, -0.18090198671715718, 0.37831814486006277, 0.08062299887788903, 6.2837707690528735, 0.301823322227404, 0.0734556127347533, 0.038060242003100875, 38.12674267246425, 0.069950983677363], 
                             [1.4504741349215766, 9.964345293695063, -0.7376990825030287, 0.05231077261958955, 0.3125101665396582, 0.2883310085055355, 6.158536076365398, 0.34929965040108363, 0.13922416746314772, 0.3762910953045031, -0.1088833839030095, 0.3674370534557028, 0.0848153341067613, 6.283365299456526, 0.3277296346681094, 0.0789335506650999, 0.041164067903278756, 37.2022470540444, 0.3727061914382968],
                             [1.4735528282275252, 10.457830123791124, -0.7378484123368936, 0.037355529918352504, 0.2547006433955337, 0.3008791197170953, 6.1422486708067865, 0.3235375223078834, 0.06800264510371007, 0.26410349603068184, -0.1617038746390121, 0.4311675329816908, 0.0760231439178584, 6.3105097093930915, 0.45160313303673505, 0.14061958460494817, -0.2384724243174104, 40.67103216913422, 0.20448222807910185],
                             [1.4453407814357084, 10.146908699569305, -0.7368584267910054, 0.01964990147236456, 0.22389466926492813, 0.2901036575777484, 6.150388714872136, 0.24173029249541889, 0.031621557419247805, 0.12143903156214009, -0.04059643374757979, 0.3656508672709845, 0.07111131914906119, 6.2705901096745595, 0.25249421901181035, 0.25049555313159316, 0.8213846079640836, 37.201743090809224, 0.3557169757535461],
                             [1.3808755265352752, 9.938472600055215, -0.7373234241292637, 0.038090695714808, 0.20848955924558116, 0.3499019345952415, 6.137333628849363, 0.20874685687808478, 0.015647339958144758, 0.18401562094411084, -0.11270699145789095, 0.5151704479477742, 0.11024561408704435, 6.207609711463972, 0.4635566755582748, 0.04336617468826924, -0.08501577607156287, 39.77586219175194, 0.2241197817497454],
                             [1.3939678263044752, 10.032643585858235, -0.7378686469911245, -0.008163823813762433, 0.18064468998438807, 0.35549933794118727, 6.158157479592278, 0.1959883903008687, 0.05891540075708436, -0.18799746415470875, -0.09159001768682726, 0.39206722407268807, 0.05405417660686076, 6.33900627520628, 0.39021235725694337, 0.037841463115627395, 0.006661094136515722, 37.83603814467343, 0.14214428958513572],
                             [1.3882997991500199, 9.530617629629507, -0.7371227044400374, 0.0339982038930266, 0.2510280096822779, 0.2858787658417807, 6.160714815084994, 0.3135805766607258, 0.03732686055281593, 0.19705566106299255, -0.12764572477705569, 0.27824366870387945, 0.07300049378676185, 6.256549164904389, 0.261062611172553, 0.030864253237076814, -0.06273794537905707, 40.3896846420874, 0.1933777121353478],
                             [1.3963648946694658, 9.75180918818749, -0.7376789256868118, 0.04657606046523677, 0.207873337295747, 0.3197216371519362, 6.155871691557279, 0.2269811867788805, 0.0673456883838116, 0.10725984535658561, -0.08261332530872764, 0.24761442522190863, 0.0659951064231678, 6.260923369555037, 0.2477448120085302, 0.011670327376688053, 0.02494415987022336, 40.511153231310246, 0.10846480606826205],
                             [1.4186702446539952, 9.953520353284203, -0.737648320356077, 0.07808284376855712, 0.18872314542308272, 0.3185394382070733, 6.162273568477383, 0.3183941063273488, 0.06616832902059701, 0.1514926906332965, -0.11957389421224227, 0.4796207326694473, 0.06235653566369398, 6.321227088580457, 0.49361428099559557, 0.10689112819233139, -0.22389994753446957, 36.6482731608843, 0.14938838767845405],
                             [1.3787010362420065, 9.361996320671075, -0.7368516812337672, 0.035454524245767005, 0.26905101733036024, 0.24193662028865287, 6.177181907420643, 0.3300817447295375, 0.07238385399488184, 0.15105870445248648, -0.09711872283124207, 0.4993698858554692, 0.059190573473550376, 6.314681233490582, 0.4593524534164035, 0.11248086924505983, 0.27511595104407505, 40.56730410570455, 0.03656999657941817],
                             [1.3580021420080053, 9.322915124718374, -0.7368940262436027, 0.05492921076747426, 0.2348390891127287, 0.2857530623058969, 6.146885871788459, 0.34170106843859654, 0.05275083240658456, 0.12918683319018243, -0.048489812906943736, 0.440667075592995, 0.11308490624596565, 6.231901481614595, 0.4176169942397346, 0.16327455479758568, 0.2525615807538353, 43.08538477871486, 0.5053992005980066],
                             [1.4641891149190578, 9.753688831797174, -0.7379197878414928, 0.10009347169472876, 0.2871607970622169, 0.25477457029082534, 6.167602616866395, 0.3199608797521209, 0.10078890745658649, 0.4656943799322155, -0.1773768732856713, 0.31327972739655935, 0.11314131130604736, 6.268509044716219, 0.3526694912213938, 0.09356834407584315, -0.036039379591684614, 46.47057812181873, 0.4128718649157406],
                             [1.4030865876511545, 10.12386582429333, -0.737301463289495, 0.060689120323593915, 0.16473204546284323, 0.3270277317527089, 6.154129856639079, 0.18337297900801677, 0.0643035953633014, 0.3941169148887802, -0.06651607914796773, 0.3863107336992121, 0.0773954202891921, 6.219636265014224, 0.47415701705156127, 0.076670263099481, 0.11962485685015434, 39.43114795339812, 0.11528196288117881],
                             [1.3908219733986031, 10.165977977664983, -0.736994799173803, 0.04626243665486435, 0.19673301181783043, 0.36558995502760305, 6.136186399671636, 0.20813324506389438, 0.07967659181739094, 0.19080401977302675, -0.14582367711583719, 0.4959518942072795, 0.06270823319393146, 6.261927632333782, 0.48955272992646537, 0.05703423717565083, -0.06418005965328531, 43.87327156688406, 0.3836933338966083]])

calculating the likelihoods as

for ray_error_vector in ray_error_vectors:
    print(main.likelihood(ray_error_vector, reinitialise=True))

for the first time prints

-37043.9913545
-36497.1396261
-39029.1026809
Warning: HotRegion.RayError raised.
Parameter vector:  [1.4735528282275252, 10.457830123791124, -0.7378484123368936, 0.037355529918352504, 0.2547006433955337, 0.3008791197170953, 6.1422486708067865, 0.3235375223078834, 0.06800264510371007, 0.26410349603068184, -0.1617038746390121, 0.4311675329816908, 0.0760231439178584, 6.3105097093930915, 0.45160313303673505, 0.14061958460494817, -0.2384724243174104, 40.67103216913422, 0.20448222807910185]
[-7.94188579e+89]
-36899.6237636
-42989.1700518
-36036.5098474
-36707.4067268
Warning: HotRegion.RayError raised.
Parameter vector:  [1.3963648946694658, 9.75180918818749, -0.7376789256868118, 0.04657606046523677, 0.207873337295747, 0.3197216371519362, 6.155871691557279, 0.2269811867788805, 0.0673456883838116, 0.10725984535658561, -0.08261332530872764, 0.24761442522190863, 0.0659951064231678, 6.260923369555037, 0.2477448120085302, 0.011670327376688053, 0.02494415987022336, 40.511153231310246, 0.10846480606826205]
[-1.18676754e+89]
-40798.7330209
-38633.244213
-37748.5032624
-39338.9084573
Warning: HotRegion.RayError raised.
Parameter vector:  [1.4030865876511545, 10.12386582429333, -0.737301463289495, 0.060689120323593915, 0.16473204546284323, 0.3270277317527089, 6.154129856639079, 0.18337297900801677, 0.0643035953633014, 0.3941169148887802, -0.06651607914796773, 0.3863107336992121, 0.0773954202891921, 6.219636265014224, 0.47415701705156127, 0.076670263099481, 0.11962485685015434, 39.43114795339812, 0.11528196288117881]
[-6.70283411e+89]
-36202.9432183

and for the second time prints

-37043.9913545
-36497.1396261
-39029.1026809
-36787.4490869
-36899.6237636
-42989.1700518
-36036.5098474
-36707.4067268
-36368.6065683
Warning: HotRegion.RayError raised.
Parameter vector:  [1.4186702446539952, 9.953520353284203, -0.737648320356077, 0.07808284376855712, 0.18872314542308272, 0.3185394382070733, 6.162273568477383, 0.3183941063273488, 0.06616832902059701, 0.1514926906332965, -0.11957389421224227, 0.4796207326694473, 0.06235653566369398, 6.321227088580457, 0.49361428099559557, 0.10689112819233139, -0.22389994753446957, 36.6482731608843, 0.14938838767845405]
[-7.73923494e+89]
-38633.244213
-37748.5032624
-39338.9084573
Warning: HotRegion.RayError raised.
Parameter vector:  [1.4030865876511545, 10.12386582429333, -0.737301463289495, 0.060689120323593915, 0.16473204546284323, 0.3270277317527089, 6.154129856639079, 0.18337297900801677, 0.0643035953633014, 0.3941169148887802, -0.06651607914796773, 0.3863107336992121, 0.0773954202891921, 6.219636265014224, 0.47415701705156127, 0.076670263099481, 0.11962485685015434, 39.43114795339812, 0.11528196288117881]
[-5.48656311e+89]
-36202.9432183

Change mention of master to main on Zenodo

Solution for showing the true parameters in corner plots with subset of all parameters.

The current version of showing cross hairs for true parameter values only works if plotting all the parameters in one corner plot (with a specific order). Here a solution (replace the existing lines with these) to make it work for any subset of parameters (in any order) shown in the corner plot (_corner.py https://github.com/xpsi-group/xpsi/blob/master/xpsi/PostProcessing/_corner.py):

if crosshairs:
    # only for topmost posterior
    self._add_crosshairs(plotter, self.params.names, self.subset_to_plot[0].truths)

...

    def _add_crosshairs(plotter, names, truths):
        """ Add parameter crosshairs to triangle plot. """
        spine = next(plotter.subplots[0,0].spines.itervalues())
        lw = spine.get_linewidth()
        for i, name in enumerate(names):        
            true_val = truths[name]
            if true_val is not None:            
                for ax in plotter.subplots[:,i]:
                    if ax is not None:
                        ax.axvline(true_val, color='black', ls='-', lw=lw)                            
                if i > 0:
                    for ax in plotter.subplots[i,:i]:
                        if ax is not None:
                            ax.axhline(true_val, color='black', ls='-', lw=lw)

Problem compile model atmosphere for "elsewhere" regions.

If I use archive/elsewhere/numerical.pyx for the elsewhere_radiation_field.pyx the compiler crashes:

/users/p19056/webb/.local/lib/python2.7/site-packages/Cython/Compiler/Main.py:369: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: /olympe3/p19056/webb/Softwares/xpsi/xpsi/surface_radiation_field/elsewhere_radiation_field.pxd
  tree = Parsing.p_module(s, pxd, full_module_name)

Error compiling Cython file:
------------------------------------------------------------
...

ctypedef struct DATA:
    size_t ndims
    size_t N[4]
    size_t BLOCKS[3]
    const srcRadField_PRELOAD *p
         ^
------------------------------------------------------------

xpsi/surface_radiation_field/elsewhere_radiation_field.pyx:39:10: 'srcRadField_PRELOAD' is not a type identifier

followed by many more Cython related errors.

Using the numerical.pyx for the hot_radiation_field works fine.
Using blackbody.pyx for the elsewhere or the hot radiation fields work fine.
So this seems to be a problem with elsewhere/numerical.pyx only.

This problem happens both on my Mac and on the cluster.

NameError raised regarding channel_edges

https://github.com/ThomasEdwardRiley/xpsi/blob/753527d46fcd6d19186d2be468d6ecc0f2304a56/xpsi/Instrument.py#L261

Bug in `module generator.py`: Wrong error type in line 1503

If no argument is passed to argparse, and then numpy.loadtxt tries to read it, then it leads to a ValueError and not IOError. The latter is raised in case we provide a wrong file name or path. A wrong path shouldn't execute the except portion of the code, but should instead raise the error and halt the program.

xpsi/xpsi/module_generator.py

Line 1503 in 2d65dd8

except IOError:

Checking possibility of allowing more than 5 `IDs` to be post-processed at the same time.

xpsi/xpsi/PostProcessing/_postprocessor.py

Line 153 in 29ead0d

if len(self._subset) > 5:

The above code and a few other following lines in the same function seem to deliberately limit the number of IDs to be post-processed to be less than or equal to 5. The objective is to allow as many as the user desires if possible.

Typos for dates in changelog

Hi , I've noticed that the latest version v[0.7.10] is dated at 2021-10-02 instead of 2022-10-02 .
Furthermore, for the 3 latest update , the date's format have switched from [YYYY-MM-DD] to [YYYY-DD-MM] , it could be better to keep only one format for the entire changelog to stay consistent.
Best regards,
Gwénaël

`KeyError: 'No parameter in subspace with matching name.'` when auto-generating CST/EST + DT models

xpsi/xpsi/module_generator.py

Line 2766 in ed9186a

if self._overlap(ref, '{0}', '{1}', 'cede', 'cede', {2:.1f}, {3:.1f}):

xpsi/xpsi/module_generator.py

Line 2815 in ed9186a

if self._overlap(ref, '{0}', '{1}', 'super', 'cede', {2:.1f}, {3:.1f}):

xpsi/xpsi/module_generator.py

Line 2819 in ed9186a

if self._overlap(ref, '{0}', '{1}', 'cede', 'cede', {2:.1f}, {3:.1f}):

The above lines check for overlap between DT super and cede regions with CST/EST cede regions. The latter is a fixed parameter for CST/EST and is therefore not available in the parameter subspace which only consists of free parameters. The check is to be performed against super region of CST/EST instead of cede.

AttributeError: 'CustomSignal' object has no attribute '_loglikelihood'

Calling the likelihood check function seem to result in AttributeError: 'CustomSignal' object has no attribute '_loglikelihood' error at least in the following cases:

If calling originally the star object with the same parameter vector p (by star(p)) as that which is provided to the likelihood check function (likelihood.check(None, [-3.27536126e+04], 1.0e-6,physical_points=[p])), and not clearing the cache (likelihood.clear_cache()) before the likelihood check or possibly not calculating the likelihood by calling the likelihood object directly first (likelihood(p)). This happens because the likelihood check then thinks that likelihood was already calculated (and the attribute _loglikelihood would then exist) even though it was actually not.
As in the previous case, but if the parameter vector provided is out of the prior bounds, calling just first likelihood(p) does not help since the likelihood will not be fully calculated and the attribute _loglikelihood will not be introduced. Clearing the cache before the likelihood check will anyway solve the problem.

I would suggest either handling this error better to give a more helpful error message about the likely cause of the issue or clearing the cache always by default before doing the rest of the likelihood check (which I think would always prevent the error from happening, at least for these cases). Clearing the cache could be automatically done at the beginning of the check() function in Likelihood.py by adding a line: self.clear_cache(). Unless people think it is good to keep the possibility for cached likelihood checks if parameters are unchanged?

The new instrumental phaseshift parameter is not merged to the parameter subspace

In the X-PSI version 0.7.9, in Signal.py the following line will produce an error (line 498) return self._shifts + self['phase_shift'] (KeyError: 'No parameter in subspace with matching name.'). The error occurs when trying to plot the signal. For fixing the error, I suggest merging this phaseshift parameter to the parameter subspace by changing (see following line in Signal.py:

xpsi/xpsi/Signal.py

Line 152 in c904dce

super(Signal, self).__init__(self._instrument,

)

# merge the subspaces; order unimportant
super(Signal, self).__init__(self._instrument,
                                   self._background,
                                   self._interstellar,
                                   *args, **kwargs)

 # merge the subspaces; order unimportant
 super(Signal, self).__init__(self._instrument,
                                     self._background,
                                     self._interstellar,
                                     phase_shift,
                                     *args, **kwargs)

Missing factor of 2 in gravitational radius

I think the function grav_radius in global_imports.py is missing a factor of 2. See also definition on wikipedia.

Missing `'old_API'` in `transform` method auto-generated in `CustomPrior.py`

xpsi/xpsi/module_generator.py

Line 3297 in dd6b51f

def transform(self, p):

The above line generates the following error message:
TypeError: transform() got an unexpected keyword argument 'old_API'

Typo in xpsi Postprocessing code

Hi , i found a typo in xpsi/PostProcessing/_pulse.py
in function " _add_true_registered_signals(self): " line 474 "eomponent" instead of what i believe should have been "component"

Update general documentation for v1.0

General documentation - Update Applications, Teams, Future, Contributing (references to xpsi-dev) and Acknowledgements.

Add something about the saved credible intervals in numerical format in the documentation as well.

Add mentions or examples in the documentation of how to obtain the credible intervals in ascii format using corner-plotter.

Error raised when phase shift is derived

https://github.com/ThomasEdwardRiley/xpsi/blob/753527d46fcd6d19186d2be468d6ecc0f2304a56/xpsi/HotRegion.py#L267

Check for more possible numerical issues when comparing 2 float numbers

Consider improving the numerical accuracy/safety by modifying the conditions where currently checking if a float has exactly a certain value (to allow some rounding errors).

Likelihood synthetise method broken

The likelihood synthesise method return none ( hence does not synthesise data) when one of the parameters is not within the bounds instead of throwing a "StrictBoundsError" error.

`a_xi` to be replaced by `a_zeta` and class attribute access to be provided using `self`

xpsi/xpsi/module_generator.py

Line 3102 in c41dea1

return x - a_xi - x*np.log(x/self.b_zeta)

Matplotlib imported in Photosphere.py even though it would not have to

The try-except-else clause checking if Matplotlib can be imported in xpsi/Photosphere.py is currently useless since it never gives any exception as _mpl is always set to None in xpsi/global_imports.py if no matplotlib can be imported and try: None does not raise any exception:

xpsi/xpsi/Photosphere.py

Line 21 in 6b7c8d8

try:

Add all the things from google doc to issues

We have some issues in a google doc, it needs to be in GitHub issues. Make sure that all issues are actually created.

update joss branch with newest version of code

Before submitting the JOSS paper, we should update the JOSS branch with the latest version of the main branch on Friday.

xpsi-group / xpsi Goto Github PK

xpsi's Introduction

X-PSI

Installation and Testing

Documentation

How to get in touch or get involved

Citing XPSI

Copyright and Licensing

Legacy

xpsi's People

Contributors

Stargazers

Watchers

Forkers

xpsi's Issues

Recommend Projects

Recommend Topics

Recommend Org