cprobability exception

[SquirrelKnight]

Please fill in the following details as much as possible when submitting your issue as this will help with testing and reproducing the issue

Environment Information

{
"version": "1.0.5",
"c_extensions present": [
"cmarkov_chain_monte_carlo",
"cprobability",
"cmoment_tensor_conversion",
"cscatangle"
],
"platform": "darwin",
"num_threads": 8,
"python version": "3.7.1 (default, Dec 14 2018, 13:28:58) \n[Clang 4.0.1 (tags/RELEASE_401/final)]",
"python version info": [
3,
7,
1,
"final",
0
],
"dependency info": {
"numpy": "1.15.4",
"scipy": "1.1.0",
"matplotlib": "3.0.2",
"cython": "0.29.2",
"sphinx": "1.8.2",
"h5py": "2.8.0"
}
}

What error you are getting

Error running cython code
Traceback (most recent call last):
File "/Users/jesse/anaconda3/lib/python3.7/site-packages/MTfit/probability/probability.py", line 871, in ln_marginalise
return cprobability.ln_marginalise(ln_pdf.astype(np.float64))
File "src/MTfit/probability/cprobability.pyx", line 130, in MTfit.probability.cprobability.ln_marginalise
ValueError: Buffer has wrong number of dimensions (expected 2, got 1)
Error running cython code
Traceback (most recent call last):
File "/Users/jesse/anaconda3/lib/python3.7/site-packages/MTfit/probability/probability.py", line 1269, in exp
return cprobability.ln_exp(self._ln_pdf)
File "src/MTfit/probability/cprobability.pyx", line 1817, in MTfit.probability.cprobability.ln_exp
ValueError: Buffer has wrong number of dimensions (expected 1, got 2)

What behaviour you would expect

To not get the errors, hopefully!

Reproduction steps

This occurs after every inversion I have attempted. I could send in the input pickle file to see if the issue can be reproduced.

[djpugh]

Hi @SquirrelKnight, thanks for this - I will have a look to see if I can reproduce it - is this blocking the run (it should try again if an error is thrown) - I think I have reproduced and it is an "expected" error - I'll update the code so that it is caught

[SquirrelKnight]

Hi @djpugh ,

It doesn't block the program, and I still seem to get results just fine.

[djpugh]

I've updated the logging to make it clearer about the errors not being important - #60 - need to fix some tests which I may have a chance to do soon, but both this error and the one in #59 should be handled more nicely