Correct monthly sampling in Energy_signature.ipynb

[saroele]

@JrtPec I found a major bug in the Energy_signature notebook and tested different ways of solving it.
Please have a look at the different generators.

I'm in favour of using the 4th and even improving it so it doesn't bug when we put diff=True because of the unit conversion. Therefore, I created a milestone: https://github.com/opengridcc/opengrid/milestone/4 to group a few improvements. #173 is handled in this commit, others to be tackled.

[coveralls]

Coverage decreased (-0.05%) to 70.088% when pulling 302522b on saroele:issue172_multivarregression into ccf995a on opengridcc:develop.

[saroele]

@JrtPec bump
I'll continue developing in this branch for now. Can you check out the bugs I solved and the solution I propose? Tnx

[coveralls]

Coverage decreased (-0.3%) to 69.446% when pulling 2993238 on saroele:issue172_multivarregression into f059fdd on opengridcc:develop.

[coveralls]

Coverage decreased (-0.3%) to 69.446% when pulling 35de4a7 on saroele:issue172_multivarregression into f059fdd on opengridcc:develop.

[saroele]

I propose to merge this branch. The multi-variable regression will still evolve and improve a lot, but no need to keep it separated from the develop branch.

[coveralls]

Coverage decreased (-0.05%) to 69.446% when pulling cf9e81c on saroele:issue172_multivarregression into 3165cda on opengridcc:develop.