We had a team meeting today, and based on the discussion, the explanation of the behavior currently observed in IDC portal is the following (to the best of my understanding):
-
Plots that correspond to the groups of facets that have selected facets show counts of cases that have the data for those selected facets.
-
Plots that correspond to the groups of facets that do not have selected facets show counts of cases for all of the facets included in the group of facets for the cohort defined by the current filter.
As such, plots can only be interpreted in conjunction and in the context of the defined filter.
So in the example below, where “CT” is selected for the “Modality” facet group, “Modality” piechart shows the number of cases that have CT Modality. The other facets show the counts for the corresponding facet groups for all of the cases that are included in the cohort defined by the filter.
@wlongabaugh @spaquett @george.white would be great if you could confirm my summary above is correct, and feel free to amend for conciseness/clarity, since we will be using this for the documentation.