The numpy boolean negative, the - operator, is not supported, use the ~ operator or the logical_not function instead.

[CristianVillenaAlemany]

I'm trying to do eLSA (expanded local similarity analysis) published in https://bitbucket.org/charade/elsa/src/master/ and the installation is fine. I did check_data and is ok. I also check all the packages necessary have greater version than required.

Name: Python Version: 2.7.18

Name: scipy Version: 1.2.2

Name: numpy Version: 1.16.5

Name: pandas Version: 0.24.2

but when I run lsa_compute I get the error:

Traceback (most recent call last): File "/home/aap/anaconda3/envs/elsa01/bin/lsa_compute", line 11, in load_entry_point('lsa==1.0.2', 'console_scripts', 'lsa_compute')() File "/home/aap/anaconda3/envs/elsa01/lib/python2.7/site-packages/lsa-1.0.2-py2.7-linux-x86_64.egg/lsa/lsa_compute.py", line 331, in main secondFactorLabels=secondFactorLabels, qvalueMethod=qvalueMethod, progressive=progressive) File "/home/aap/anaconda3/envs/elsa01/lib/python2.7/site-packages/lsa-1.0.2-py2.7-linux-x86_64.egg/lsa/lsalib.py", line 1044, in applyAnalysis trendThresh, True) #now allowing trend analysis in singleLSA File "/home/aap/anaconda3/envs/elsa01/lib/python2.7/site-packages/lsa-1.0.2-py2.7-linux-x86_64.egg/lsa/lsalib.py", line 217, in singleLSA xSeries = zNormalize(fTransform(series1)) File "/home/aap/anaconda3/envs/elsa01/lib/python2.7/site-packages/lsa-1.0.2-py2.7-linux-x86_64.egg/lsa/lsalib.py", line 804, in robustZNormalize ranks = tied_rank(tseries) File "/home/aap/anaconda3/envs/elsa01/lib/python2.7/site-packages/lsa-1.0.2-py2.7-linux-x86_64.egg/lsa/lsalib.py", line 703, in tied_rank V = V[-V.mask] #remove nan's TypeError: The numpy boolean negative, the - operator, is not supported, use the ~ operator or the logical_not function instead.

I already tried upgrading h5py but looks it is not the problem. I also tried to create a conda environment with specific packages but still does not work.

Do you have a clue of what is going on?

Thanks

Cristian