VEuPathDB / EdaNewIssues

0 stars 0 forks source link

EDA : ITN bednet count may have incorrect metadata? #616

Open moontrip opened 1 year ago

moontrip commented 1 year ago

@danicahelb (and @bobular as you are familiar with vocabulary-related works)

I am working on range and colormap for continuous overlay variable, so I was browsing variables to test gradient colormaps. For Amazonia ICEMR Brazil Cohort data, I tried to use ITN bednet count as a continuous overlay variable at Scatterplot Viz and I expected gradient colormap legend. However, it did not even show plot itself: screenshot is attached below and a shared link is provided https://clinepidb.org/analysis/UbwlOI0.

I looked into data and it seems to have correct format with seriesGradientColorscale from the backend. Thus, I checked our implementation (programming codes) and variable's metadata, and found that it has unnecessary (?) vocabulary, which makes the frontend mishandle it. I also checked other variable like GEMS1 BMI-for-age-z-score, which is also a continuous overlay variable, and it did not have the vocabulary and work fine. For comparison purpose, I have attached those variable metadata as screenshots.

Honestly, I am not quite sure if it is an issue of the metadata or frontend needs to consider such case(s). What do you think? (FYI, if the ITN bednet count is used as either xAxisVariable or yAxisVariable, then there is no problem)

a) BMI-for-age-z-score: does not have vocabulary GEMS1 BMI-for-age-z-score

b) ITN bednet count: has vocabulary Amazonia Brazil ITN bednet count