Viewer: Cohort Overlaps error

[javier-gracia-tabuenca-tuni]

[siirtola]

It might have something to do how you start the shiny application:

https://stackoverflow.com/questions/74245044/error-c-stack-usage-15924224-is-too-close-to-the-limit

[siirtola]

This is perhaps the best explation, but I still don't understand what would use so much stack space? Also, why it works in the test environment but not in Docker? The stack limit is the same.

https://stackoverflow.com/questions/14719349/error-c-stack-usage-is-too-close-to-the-limit?rq=4

[javier-gracia-tabuenca-tuni]

i think what use so much memory is this UpSetR::fromExpression seems not very efficient to spread a matrix of few count values in a binary matrix with as many values as counts to later count them and plot the counts

[siirtola]

I don't get it (first time I looked that function): why it is forming that matrix when all that is needed is in the expression? Why the re-count?

[siirtola]

Can you describe the input that caused the error? I can't get it to fail with any mock data that I have, the largest cohort being over 500,000 patients.

[javier-gracia-tabuenca-tuni]

solved, It was some flag on the docker file remaining form an other problem i tried to solve with Java running out of memory i found it bcs it was working well when I ran it inside rstudio in sandbox

[siirtola]

Good! I am still wondering why the fromExpression builds the matrix? Is it a matter of trust – are they checking the expression does not contain conflicts?