[R-meta] catching errors in batch meta-analyses
I find the functions safely() and possibly() from the purrr package for this sort of error handling stuff. For instance, create safe_reitsma <- safely(mada::reitsma) Then run X_mada <- safe_reitsma(data=subset(X,SUBMA_num==iii), correction.control="single", correction=0.5) The result will always have two components, X_mada$result and X_mada$error. Further details and examples here: https://purrr.tidyverse.org/reference/safely.html James On Sun, Aug 15, 2021 at 5:57 PM Philippe Tadger <philippetadger at gmail.com> wrote:
Dear colleges
In a batch of a several MA (simulation kind) I would like to capture the
error messages (like non-convergence) from functions like
mada::reitsma. I have been exploring the function tryCatch but without
been able to capture the error messages. Something like this
X.mada<-vector("list",745)
for (iii in 1:200) {
X.mada[iii]<-tryCatch(mada::reitsma(data=subset(X,SUBMA_num==iii),
correction.control="single", correction=0.5))
}
Ideally X.mada will store convergence and non-convergence messages.
Thanks in advance for the support!
--
Kind regards
*Philippe Tadger*
ORCID <https://orcid.org/0000-0002-1453-4105>, Reseach Gate
<https://www.researchgate.net/profile/Philippe-Tadger>
[[alternative HTML version deleted]]
_______________________________________________ R-sig-meta-analysis mailing list R-sig-meta-analysis at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis