When performing RandomSampling using RandomUnderSampler, RandomOverSampler or RandomSampler from random.py we set a desired class distribution,
but in an online setting we know we can't be 100% sure the sampling will give us the exact distribution we wanted,
so a variable to track that might be useful,
as we have _actual_dist to keep track of all the data that went through the model, I believe a _trained_on_dist might also be useful, to track the distribution of the data that was used to train the base model with the sampling technique chosen
When performing RandomSampling using RandomUnderSampler, RandomOverSampler or RandomSampler from random.py we set a desired class distribution,
but in an online setting we know we can't be 100% sure the sampling will give us the exact distribution we wanted,
so a variable to track that might be useful, as we have
_actual_dist
to keep track of all the data that went through the model, I believe a_trained_on_dist
might also be useful, to track the distribution of the data that was used to train the base model with the sampling technique chosen