Closed gowthamrao closed 8 months ago
The object results$timeRef does not have NA
but, covariates does
base R syntax gives an error on Andromeda object. pulling data out of Andromeda using dplyr::collect() resolves it. However, it takes long time -- and defeats the purpose of Andromeda
What is bothering me is - this has been working for a long time. timeId is NULL for covariates like following (that are not time dependent). i.e. we had timeId = NA as part of FeatureExtraction output for a long time and this syntax used to handle it.
https://github.com/OHDSI/CohortDiagnostics/blame/76911f757b9db74f7fe1c4a7e7405d4d285dc622/R/CohortCharacterizationDiagnostics.R#L156-L158
I got an error while running CD.
in debug mode, it was at
I think an Andromeda object (i.e. output of Feature Extraction) does not support this?