add mad wrapper

[OriolAbril]

Add a wrapper for the median_abs_deviation function in scipy.stats.

[codecov]

Codecov Report

Merging #4 (679b723) into main (6b39cc7) will decrease coverage by 0.17%.
The diff coverage is 87.50%.

@@            Coverage Diff             @@
##             main       #4      +/-   ##
==========================================
- Coverage   93.16%   92.99%   -0.18%     
==========================================
  Files           6        6              
  Lines         512      528      +16     
==========================================
+ Hits          477      491      +14     
- Misses         35       37       +2     

Impacted Files	Coverage Δ
src/xarray_einstats/stats.py	92.04% <87.50%> (-0.46%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 6b39cc7...679b723. Read the comment docs.

[canyon289]

Thanks :)

[OriolAbril]

I will "need" this next week to test that arviz-devs/arviz#1974 correctly computes the mcse for the mad correctly. If there are summary stats or functions that are missing open an issue or send a PR.

logsumexp from scipy.special should probably be added for example. Note that most special functions don't need a wrapper though, they work on dataarrays straight away because they are pointwise: softmax, expit, logit... already work on xarray objects without any wrapper needed.