OHDSI Home | Forums | Wiki | Github

Must we work with JDBC and R?

Perhaps there is a firewall between my server and the workstation, I have a lot of problems dealing with JDBC that is common in most of the OHDSI packages.

including

  • Silent Failures
  • Silent Timeouts
  • An I/O error occurred while sending to the backend.
  • SQL translation issues (with SQLrender, but that is a problem when one trys to have a generalist approach).

I have a lot of struggles with the packages and unfortunately there isn’t a lot of solutions online aka googling.

When I extract those said SQL to python, and connect with psycopg2, most of them would work.

I understand that the use of R could be a historical reason. But do we have to use R? Just curious.

I know this is unhelpful, but I’ve found that in general JDBC works really well, with hardly any issues. I’m sorry you’re experiencing issues in your particular environment.

You don’t have to use R, it is just that almost all of the OHDSI tools are written in R. Rewriting them in Python would take several person-decades.

t