robince / pyentropy Goto Github PK

Hello,

Today I have just installed your package but I came across the following error: when running the 'calculate_entropies' method I bump into:

/Users/GiuseppePietro/Library/Python/2.7/lib/python/site-packages/pyentropy/statk/init.py in ()
1
----> 2 from wrap import nsb_entropy, bub_entropy

ImportError: No module named wrap

Do you know what might be the cause?

Thanks in advance!

Giuseppe

Hi, Can you tell me what functions does this repo provide? Is there a tutorial that i can check or do you know a more comprehensive library for neural analysis in python?

Hello,

I am trying to calculate the shuffled MI estimate with the following code:

sort_sys.calculate_entropies(method='pt',calc=['HX','HXY','HiXY','HshXY'])
I_pt[ii,ts] = sort_sys.Ish()

Where sort_sys has been initialised
pye.SortedDiscreteSystem(R_ord,R_dims,S,Ns)
S is an alphabet of stimuli, and R_ord are the responses ordered by stimulus appearance as in S, R_dims and Ns are formatted as well following the paper.

I can calculate the plugin and pt estimation without problem, but the shuffled estimators are exactly the same as the non shuffled, HshXY and HiXY are equal to HXY, but the responses are being shuffled as the returned X, Xsh are different. Do you have any idea of what might be going wrong?

Thanks in advance,

Giuseppe

P.S. if I initialise a new system where the responses are shuffled beforehand, this HXY is correctly bigger than the one obtained by the non-shuffled system.

I try to use pyentropy on Windows and "python setup.py install" gives the following warning:
WARNING: The gsl-based extension module could not be compiled.
pyEntropy should run, but the features present in that file will not be available.

I see you are using Windows. The default compiler for this platform is the Microsoft Visual
Studio C compiler. However, a free alternative compiler called mingw can be used instead.

I have the right compiler installer (VS2008) and could use it to compile other Python extensions.

• Should this work? Is this a bug?
• What functionality is lost due to this?
  Thanks!

Hi, sample_entropy = ent.sample_entropy(ts, 4, 0.2 * std_ts) in this method what is the significance of 4 , what it stands for?