[MRG] Add DomainAndLabelStratifiedSubsampleTransformer + Fix DomainStratifiedSubsampleTransformer

[YanisLalou]

• Fix TypeError: DomainStratifiedSubsampleTransformer.transform() got an unexpected keyword argument 'allow_source'with DomainStratifiedSubsampleTransformer
• Add DomainAndLabelStratifiedSubsampleTransformer to have a transformer preserving the percentage of samples
  for each class and for each sample domain (To prevent issues like: ValueError: This solver needs samples of at least 2 classes in the data, but the data contains only one class: 1)

[antoinecollas]

Then, we have to be consistent with the splitters. IMO, we have to rename the subsamplers as:

• DomainStratifiedSubsampleTransformer --> DomainSubsampleTransformer
• DomainAndLabelStratifiedSubsampleTransformer --> StratifiedDomainSubsampleTransformer

[antoinecollas]

BTW, we can even get rid of 'Transformer':

• SubsampleTransformer --> Subsampler
• DomainStratifiedSubsampleTransformer --> DomainSubsampler
• DomainAndLabelStratifiedSubsampleTransformer --> StratifiedDomainSubsampler

[YanisLalou]

I agree