Open tafsiri opened 8 years ago
I'm still in favour of leaving this all in the analysis code. At least for V1. It might be that in some future iteration visualizations might provide an api for filtering out additional words from display in cases where that wouldn't distort the visualization and underlying data.
However structurally we may want to provide a mechanism for visualization to incorporate a chain (abstractly) of pre-processing functions to their input data (which would initially be empty—but could eventually be a stopword filter or something similar)
@arnicas ping!
As I said, if it's super easy to filter a new stopword out and get to the vis update, ok with leaving it in analysis. I think the config object should have the stopwords lists as a part of it, for potential onscreen display as we discussed. Yes, I'll update that issue tool.
We discussed this a fair amount and seemed to land on moving this all to the analysis scripts. However it came up again in our internal kickoff/review, so this issue is to track discussion around that.