[ENH] New stats functions

[liuzhenqi77]

Adding some new stats functions. Most of them are numba-optimized, but also works without numba.

They'll ultimately be part of Andrea's toolbox. I'm putting the alpha versions here so we can test them & be early adopters.

• network_pearsonr
• network_pearsonr_numba
• network_pearsonr_pairwise
• effective_resistance
• network_polarisation
• network_variance
• network_variance_numba
• network_covariance
• network_covariance_numba

[codecov-commenter]

Codecov Report

Merging #126 (730c230) into master (4b6e954) will decrease coverage by 2.23%. Report is 1 commits behind head on master. The diff coverage is 20.42%.

:exclamation: Current head 730c230 differs from pull request most recent head 2add592. Consider uploading reports for the commit 2add592 to get more accurate results

:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.

@@            Coverage Diff             @@
##           master     #126      +/-   ##
==========================================
- Coverage   57.93%   55.70%   -2.23%     
==========================================
  Files          26       26              
  Lines        2249     2391     +142     
==========================================
+ Hits         1303     1332      +29     
- Misses        946     1059     +113     

Files	Coverage Δ
netneurotools/plotting.py	32.84% <ø> (ø)
netneurotools/metrics.py	15.16% <40.00%> (+0.45%)	:arrow_up:
netneurotools/stats.py	48.21% <19.70%> (-15.32%)	:arrow_down:

[liuzhenqi77]

I'm merging this for now, however, more testing is needed.