Add choice of estimation weigths and standard deviation for RandomForestRegressor models

[HealthyPear]

This PR adds the key estimation_weight in analysis.yaml for both estimators.

For the energy, the choice can be either the standard deviation from the single forest's trees or an evaluation of a DL1 basic feature like for the model building.

For the classifier, the choice is also an evaluation of basic DL1 quantities.

[codecov]

Codecov Report

Merging #134 (24c15a9) into master (ff7ea05) will decrease coverage by 0.09%.
The diff coverage is 64.00%.

@@            Coverage Diff             @@
##           master     #134      +/-   ##
==========================================
- Coverage   60.46%   60.36%   -0.10%     
==========================================
  Files          23       23              
  Lines        2102     2122      +20     
==========================================
+ Hits         1271     1281      +10     
- Misses        831      841      +10     

Impacted Files	Coverage Δ
protopipe/scripts/data_training.py	92.75% <54.54%> (-2.20%)	⬇️
protopipe/scripts/write_dl2.py	81.41% <71.42%> (-1.04%)	⬇️

---

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 ff7ea05...24c15a9. Read the comment docs.