Closed mdlavallee92 closed 1 month ago
Thanks @mdlavallee92 - appreciate the feedback! I've faced the same issues regarding ParallelLogger and RMarkdown.
I like the idea of having more control over the messaging as you have proposed. I've also considered removing the use of ParallelLogger for messaging and just using the native message functions. I'll put this into ideas for the next CG release.
Cool! Let me know how I can help!
@mdlavallee92 - I've made some changes on the develop
branch that should address this and will be part of the v0.10 release. I'll close this for now but if you face any problems let me know and I'll re-open. Thanks!
First off, love
CohortGenerator
, it is super handy in studies! Nice work @anthonysenaI am trying to use
CohortGenerator
in markdown chunks in a way to run interactive OHDSI pipelines. However a problem I have is thatParallelLogger
conflicts with theglobalCallingHandlers
of the markdown chunk. This is a known problem for any logger package in markdown and is not a bug in either package. To avoid this error when running markdown chunks, I currently wrap theCohortGenerator
functions inpurrr::quietly
to collect the output as a message.Is it within the development roadmap to add a feature to the
CohortGenerator
functions that offers a toggle, such asverbose = TRUE
, that can turn on or off theParallelLogger
within the function?For example: