Open njr2128 opened 2 years ago
Even just doing percentages might help
Another way would be to calculate average/mean of tag of interest across ALL categories and then look at deviations in each category
A way to think about or explain this: what is the probability that a certain tag will occur in a chosen category?
Visualizations like the number of deletions and additions by category are really useful, but it would be really interesting to see normalize the data to see whether they are over/underrepresented in a certain category. Obvivously, there are many adds/dels in casting, but that is bc there are more casting entries than any other category. If we were to normalize this, are there any categories that have more dels/adds than others?
Can we do this for many of the charts in https://cu-mkp.github.io/sandbox/docs/Kaufman_final-report.html?