unclear documentation at RandomUndersampler and RandomOverSampler

[snowparrot]

Hello, there are two minor issues I see in the docs of RandomUnderSampler and RandomOversampler.

• There is the constraint that the float in sampling_strategy should be between 0 and 1. It should be documented explicitly.
• . "Therefore, the ratio is expressed as \alpha_{us} = N_{rM} / N_{m} where N_{rM} and N_{m} are the number of samples in the majority class after resampling and the number of samples in the minority class, respectively."
  should be replaced by
  "Therefore, the ratio is expressed as \alpha_{us} = N_{rM} / N_{m} where N_{rM} is the number of samples in the majority class after resampling and N_{m} is the number of samples in the minority class after resampling , respectively."
  The new version puts the definition next to variable (clearity) and adds another "after resampling". Im a big fan of clear definition for math formulars. Also you might assume that the reader isnt a profane reader of English. Clear definition are easier to read.

[snowparrot]

oh thank you for review!

[glemaitre]

Actually, you should remove the respectively if we go with your definition. Apart of that you are welcome to open a PR, those changes will improve the documentation.
Also we should modify the definition for the over-sampling as well.

[arielrossanigo]

Hi, I guess that there are another mistake in this section of the doc regarding the formula expressed there.

In the code , the formula being used is N_{rM} = N_{m} / \alpha_{us}; so in the doc it should be \alpha_{us} = N_{m} / N_{rM}

Please, let me know if I'm interpreting something wrong.