warrenweckesser / analyze-scipy-code

Assorted tools used to analyze SciPy's source code

Assorted scripts that I use to analyze the SciPy source code.

• dists_that_override.py: Display which of the SciPy univariate continuous distributions override the given method names. For example,

  $ python dists_that_override.py _cdf _ppf _sf _isf _rvs
  SciPy version 1.11.0.dev0+1566.d5ceb77
  
  distribution         _cdf _ppf _sf  _isf _rvs
  alpha                ✔    ✔    -    -    -
  anglit               ✔    ✔    -    -    -
  arcsine              ✔    ✔    -    -    -
  argus                ✔    -    ✔    -    ✔
  beta                 ✔    ✔    ✔    ✔    ✔
  betaprime            ✔    ✔    -    -    ✔
  bradford             ✔    ✔    -    -    -
  burr                 ✔    ✔    ✔    -    -
  [...]
  wald                 ✔    ✔    ✔    ✔    ✔
  weibull_max          ✔    ✔    ✔    -    -
  weibull_min          ✔    ✔    ✔    -    -
  wrapcauchy           ✔    ✔    -    -    -
  

• find_functions_missing_examples.py: Find functions whose docstring is missing the "Examples" section.

• find_missing_import_np.py: Find functions where there is an "Examples" section that uses the name np but that do not have a corresponding import numpy as np.

• find_docstring_issues.py: (Work in progress) Looks for several common issues in the docstrings. For example,

  $ python find_docstring_issues.py sparse
  scipy version 1.10.0.dev0+2299.5beb395
  
  === sparse ===
  sparse.block_diag
      section out of order: 'See Also'
  sparse.diags
      missing section: 'Returns'
  sparse.eye
      missing section: 'Returns'
  sparse.hstack
      missing section: 'Returns'
  sparse.identity
      missing section: 'Returns'
  sparse.rand
      section out of order: 'See Also'
  sparse.random
      duplicated imports in Examples:
          >>> from scipy.sparse import random
  sparse.save_npz
      missing section: 'Returns'
  sparse.spdiags
      missing section: 'Returns'
  sparse.vstack
      missing section: 'Returns'
  

• extract_example_code.py: Extract the code from the 'Examples' section of the SciPy object.

  For example,

  $ python extract_example_code.py scipy.special.logsumexp
  Code from the 'Examples' section of scipy.special.logsumexp written to 'example_logsumexp.py'.
  $ cat example_logsumexp.py
  # Python code extracted from the 'Examples' section of
  # scipy.special.logsumexp
  
  import numpy as np
  from scipy.special import logsumexp
  a = np.arange(10)
  logsumexp(a)
  np.log(np.sum(np.exp(a)))
  a = np.arange(10)
  b = np.arange(10, 0, -1)
  logsumexp(a, b=b)
  np.log(np.sum(b*np.exp(a)))
  logsumexp([1,2],b=[1,-1],return_sign=True)
  a = np.ma.array([np.log(2), 2, np.log(3)],
                  mask=[False, True, False])
  b = (~a.mask).astype(int)
  logsumexp(a.data, b=b), np.log(5)