numpy math or emath

[arefpz]

Hi,

I just started using the toolbox to find effective connectivity in EEG data. I followed all the instructions on the GitHub page to install it in the conda environment. First, there was an issue with "statsmodels" that Python couldn't import; I then installed it using conda, and the issue was resolved. However, I then received this error when trying to run the demo examples.
It looks like the error is AttributeError: module 'numpy' has no attribute 'math'. Did you mean: 'emath'?

file "/home/aref/Documents/IDTxl-1.6.0/demos/demo_multivariate_mi.py", line 21, in <module> results = network_analysis.analyse_network(settings=settings, data=data) File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/idtxl/multivariate_mi.py", line 161, in analyse_network res_single = self.analyse_single_target( settings, data, targets[t], sources[t] ) File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/idtxl/multivariate_mi.py", line 288, in analyse_single_target self._include_source_candidates(data) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^ File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/idtxl/network_inference.py", line 974, in _include_source_candidates self._include_candidates(candidates, data) ~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^ File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/idtxl/network_inference.py", line 151, in _include_candidates significant = stats.max_statistic( ~~~~~~~~~~~~~~~~~~~^ self, ^^^^^ ...<3 lines>... conditional=self._selected_vars_realisations, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ )[0] ^ File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/idtxl/stats.py", line 458, in max_statistic _check_permute_in_time(analysis_setup, data, n_perm) ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/idtxl/stats.py", line 1669, in _check_permute_in_time if not analysis_setup.settings["permute_in_time"] and not _sufficient_replications( ~~~~~~~~~~~~~~~~~~~~~~~~^ data, n_perm ^^^^^^^^^^^^ ): ^ File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/idtxl/stats.py", line 1535, in _sufficient_replications if np.math.factorial(data.n_replications) > n_perm: ^^^^^^^ File "/home/aref/anaconda3/envs/idtxl/lib/python3.13/site-packages/numpy/__init__.py", line 414, in __getattr__ raise AttributeError("module {!r} has no attribute " "{!r}".format(__name__, attr)) AttributeError: module 'numpy' has no attribute 'math'. Did you mean: 'emath'?

I am using Ubuntu 22.04.5
Could you please help me figure out what I need to do to fix this error?
Thanks,

[arefpz]

Okay. I solved the problem by installing statsmodels
conda install statsmodel
and replacing the np.math.factorial with factorial and adding from math import factorial in idtxl/stats.py.
You will need to install the idtxl after the modification if you have changed it in the pulled repository.
At some point, I also needed to install Java, since my Linux system was fresh.
I suggest adding these to the installation instructions.
Thanks.

[felixkoerber]

Alternatively, you can change the np.math.factorial to sc.special.factorial and add "import scipy as sc" at the start of the script, where the error occurs.