OHDSI Home | Forums | Wiki | Github

Avoiding Dataframes in AWS Glue


Does anyone have experience implementing the CDM in AWS?

I have my infrastructure built and am ready to execute a glue job against my source data to transform it into the CDM, and my organization wants to avoid creating a dynamic dataframe in the Glue job that will do the transformation. Most examples of implementation use CASE WHEN logic from SQL but it is when I include a SQL Query node in my Glue job that Glue generates the dynamic dataframe.

Any advice in welcome.