As found in this answer, your best bet is to install numpydoc
and its requirements.
import numpydoc
import numpy as np
doc = numpydoc.docscrape.NumpyDocString(np.random.binomial.__doc__)
wich can then be inspected
In [45]: doc['Parameters']
Out[45]:
[('n',
'int or array_like of ints',
['Parameter of the distribution, >= 0. Floats are also accepted,',
'but they will be truncated to integers.']),
('p',
'float or array_like of floats',
['Parameter of the distribution, >= 0 and <=1.']),
('size',
'int or tuple of ints, optional',
['Output shape. If the given shape is, e.g., ``(m, n, k)``, then',
'``m * n * k`` samples are drawn. If size is ``None`` (default),',
'a single value is returned if ``n`` and ``p`` are both scalars.',
'Otherwise, ``np.broadcast(n, p).size`` samples are drawn.'])]
Note that you'll have to do some postprocessing such as converting the list of tuples to a dictionary.
In [46]: {t[0]: t[1] for t in doc['Parameters']}
Out[46]:
{'n': 'int or array_like of ints',
'p': 'float or array_like of floats',
'size': 'int or tuple of ints, optional'}