Add MeanStdScaling transform

[bencottier]

Closes #3

Note: the notion of dims is now consistent with mapslices and Statistics.mean/std rather than eachslice. To do that, I inverted the dims passed to the general apply! method for arrays. Statistics.mean/std was used in FeatureEngineering. Related issue: #18

[codecov]

Codecov Report

Merging #17 (e8970b0) into main (9fd201a) will increase coverage by 4.58%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main      #17      +/-   ##
==========================================
+ Coverage   94.50%   99.09%   +4.58%     
==========================================
  Files           8        9       +1     
  Lines          91      110      +19     
==========================================
+ Hits           86      109      +23     
+ Misses          5        1       -4     

Impacted Files	Coverage Δ
src/Transforms.jl	100.00% <ø> (ø)
src/scaling.jl	100.00% <100.00%> (ø)
src/temporal.jl	100.00% <100.00%> (ø)
src/transformers.jl	100.00% <100.00%> (+4.16%)	⬆️
src/utils.jl	87.50% <100.00%> (+87.50%)	⬆️
src/power.jl	100.00% <0.00%> (ø)
src/periodic.jl	100.00% <0.00%> (ø)
src/one_hot_encoding.jl	100.00% <0.00%> (ø)
src/linear_combination.jl	100.00% <0.00%> (ø)
... and 2 more

---

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 9fd201a...8730f3a. Read the comment docs.

[nicoleepp]

I noticed you've been merging branches into this one, can you try rebasing instead in the future? It keeps the git history much more clean.

[bencottier]

I have done some refactoring to always populate the mean and std at construction time (either precomputed or computed). I added the name keyword argument to _apply! inside the general transformers.jl methods for compatibility.

Overall it seems better - much less overloading. One thing is dims or cols has to be specified both at construction and apply if used.

[nicoleepp]

Note: the notion of dims is consistent with eachslice. Statistics.mean/std and mapslices have the opposite notion of dims (e.g. dims=1 ~ dims=2 and dims=3 ~ dims=[1, 2]). Statistics.mean/std was used in FeatureEngineering. Related issue: #18

Why did you decide to go with dims being consistent with eachslice instead of what we are already doing in FeatureEngineering?

[nicoleepp]

Overall looking a lot cleaner now I think

[bencottier]

Note: the notion of dims is consistent with eachslice. Statistics.mean/std and mapslices have the opposite notion of dims (e.g. dims=1 ~ dims=2 and dims=3 ~ dims=[1, 2]). Statistics.mean/std was used in FeatureEngineering. Related issue: #18

Why did you decide to go with dims being consistent with eachslice instead of what we are already doing in FeatureEngineering?

Because the other transforms so far use dims consistent with eachslice. But I think this is the first transform where dims actually changes the results. So given that, and the FeatureEngineering convention, I'm now leaning towards switching over to mapslices and its convention. Shall I go ahead with that?

[nicoleepp]

Note: the notion of dims is consistent with eachslice. Statistics.mean/std and mapslices have the opposite notion of dims (e.g. dims=1 ~ dims=2 and dims=3 ~ dims=[1, 2]). Statistics.mean/std was used in FeatureEngineering. Related issue: #18

Why did you decide to go with dims being consistent with eachslice instead of what we are already doing in FeatureEngineering?

Because the other transforms so far use dims consistent with eachslice. But I think this is the first transform where dims actually changes the results. So given that, and the FeatureEngineering convention, I'm now leaning towards switching over to mapslices and its convention. Shall I go ahead with that?

The other mutating transforms follow eachslice, but the non mutating follow the mapslices convention. So yeah I think we should move everything to follow the mapslices convention for dims

[bencottier]

Sorry, I made a mess of commits trying to rebase after #19 . I will squash before merging.