Replace multiprocessing with joblib, parallelize ESS and TLE

[JesseCresswell]

This PR addresses issue #21
I used this test code to demonstrate the issue and benchmark the solution:

import numpy as np
import skdim
import time

np.random.seed(0)
X = np.random.random((200000, 10))

estimator = skdim.id.ESS() # LocalEstimator
# estimator = skdim.id.FisherS() # GlobalEstimator
# estimator = skdim.id.lPCA() # Special Case
n_jobs = 1

t0 = time.time()
if isinstance(estimator, skdim._commonfuncs.LocalEstimator):
    lid = estimator.fit(X, n_neighbors=10, n_jobs=n_jobs).dimension_pw_
elif isinstance(estimator, skdim._commonfuncs.GlobalEstimator):
    lid = estimator.fit_pw(X, n_neighbors=10, n_jobs=n_jobs).dimension_pw_
t1 = time.time()

print(lid.mean())
print(t1-t0)

LocalEstimator ESS
On commit 427b1881b before my changes, I time the kNN computation and the ESS computation separately and vary n_jobs. I observe the following using a computer with 40 cpu.

n_jobs	Time kNN (s)	Time ESS (s)	Mean LID
1	92.71	146.98	8.959729
2	91.05	149.36	8.959729

After the PR changes I observe

n_jobs	Time kNN (s)	Time ESS (s)	Mean LID
1	95.44	150.98	8.959729
2	79.22	91.84	8.959729
20	99.55	17.03	8.959729
40	70.97	18.94	8.959729

Conclusions:
The ESS estimator code can be parallelized effectively. The calls to kNN computation have a lot of variance in terms of timing, and are not clearly benefiting from n_jobs>1

GlobalEstimator LPCA
Before these changes

n_jobs	Time kNN (s)	Time LPCA (s)	Mean LID
1	94.59	83.54	7.006095
2	91.35	69.74	7.006095
20	79.12	43.31	7.006095
40	57.86	41.53	7.006095

With this PR

n_jobs	Time kNN (s)	Time LPCA (s)	Mean LID
1	93.75	82.25	7.006095
2	88.16	47.64	7.006095
20	78.79	12.94	7.006095
40	60.44	15.94	7.006095

Conclusions:
PCA already used parallelism through the GlobalEstimator code, but joblib provides an additional speedup over multiprocessing

Additional results on TLE which I added parallelism for, and FisherS which is a GlobalEstimator that has strange parallel behaviour already, are provided in this google doc
https://docs.google.com/document/d/1Jo9VPFQwBaDP9Xj-8xiA_ueULNstXqF554VfFr6Wmeo/edit?usp=sharing

[j-bac]

thanks Jesse. Really nice speedup!

[JesseCresswell]

Those were all the changes I planned to make. It would be great if you could make a new release as I'm actively using the repo for research.

[j-bac]

Glad to hear it's useful. Just created a new pypi release