search

OHDSI / OhdsiShinyModules

An R package containing Shiny modules used by various OHDSI Shiny apps
https://ohdsi.github.io/OhdsiShinyModules/
8 stars 11 forks source link

General: Round numbers to 2 decimal places unless otherwise specified #112

Open nhall6 opened 1 year ago

jreps commented 1 month ago

Estimation CM/SCCS tables have long decimal places or to 4dp. Prediction main table has 3dp Characterization has results to 3dp Cohort Incidence is 4dp Cohort Diagnostics does % to 1pd.

We need standardization but I'm unsure whether 2pd is what we should use. Perhaps we should default results to 3dp or 4dp but make percentages to 1dp?

nhall6 commented 1 month ago

Estimation CM/SCCS tables have long decimal places or to 4dp. Prediction main table has 3dp Characterization has results to 3dp Cohort Incidence is 4dp Cohort Diagnostics does % to 1pd.

We need standardization but I'm unsure whether 2pd is what we should use. Perhaps we should default results to 3dp or 4dp but make percentages to 1dp?

I was also looking at this issue and thinking about this. I think 4 decimal places is probably best. I would vote percentages should go to 2dp. Should we check with Patrick? @pbr6cornell