Counting in review plot seems off by a lot

[ACharbonneau]

I understand that we still have the double counting issue, and that's fine. But this seems really weird. This data set only has 5 subjects.

[karlcz]

I think this is multi-counting because of the pre-processed level1_stats reflecting combinations of single terms from all different vocab "dimensions" at once. There are actually three core_fact combinations for 3 subjects that combine multiple terms on multiple vocabs, and these would expand into many single-term-per-vocab combinations.

• https://app-staging.nih-cfde.org/chaise/record/#162/CFDE:core_fact/nid=159
• https://app-staging.nih-cfde.org/chaise/record/#162/CFDE:core_fact/nid=160
• https://app-staging.nih-cfde.org/chaise/record/#162/CFDE:core_fact/nid=161

[ACharbonneau]

KK. I'll stand down then :)