Adds weighted permutation entropy

[samgdotson]

This pull request:

1. Implements weighted permutation entropy according to Fadlallah et al. (2013)
2. Adds corresponding, passing, tests
3. Changes the name of CAHNGELOG.md --> CHANGELOG.md

There is already a pull request for weighted permutation entropy, however, this PR differs in the following ways

1. Includes the time delay (tau) which is an important part of delay coordinate embedding (the previous PR neglects this).
2. Includes appropriate tests (the previous PR neglects this).
3. This implementation does not use a hashmap.

The time performance for this implementation is good (not perfect), but can still calculate PE for an arbitrarily high embedding dimension.

Time performance test:

%%timeit
time = np.linspace(0, 100, 10000)
data = np.cos(time)
m = 20
tau = 2
hw = wpe(data, order=m, delay=tau, normalize=True)
hw

>>> 148 ms ± 906 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)

[codecov]

Codecov Report

Merging #19 (0aa4c37) into master (704e919) will increase coverage by 1.01%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master      #19      +/-   ##
==========================================
+ Coverage   91.79%   92.81%   +1.01%     
==========================================
  Files           2        2              
  Lines         134      153      +19     
==========================================
+ Hits          123      142      +19     
  Misses         11       11              

Impacted Files	Coverage Δ
pyentrp/entropy.py	90.26% <100.00%> (+1.48%)	⬆️
tests/test_entropy.py	100.00% <100.00%> (ø)

---

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 704e919...0aa4c37. Read the comment docs.

[nikdon]

Hey. Sorry for the delay in communication. Thank you for the PR. I'll merge it and prepare a new release shortly