Move VOCAB size setting to parameter dictionary
tukeyclothespin opened this issue · comments
VOCAB size is currently duplicated in the BOW and LDA dictionaries. Moving it to the parameter dictionary will standardize vocabulary size across featurizers.
It is possible that someone would want different-sized dictionaries for these two techniques but I think the logic for vocab generation doesn't take this into account either. I'm okay merging both into one VOCAB param since it is more in keeping with our only-do-things-once spirit.