Thomas White

@chrisknoll , per discussion, removed Achilles analysis # 40

Thanks for the patch. Unfortunately, it is very slow on Spark. Using our code in production (which uses the logic from my patch), it takes 3 seconds to run analysis...

Proactively, I expect one concern would be that many OMOP instances do not use datetime, so use of cohorts that truly require datatime would not work on those instances. This...

@alex-odysseus or @chrisknoll , can you provide a status update on this feature request?

@chrisknoll , although I hear your concern, this capability would be high value for EHR data, and electronic clinical quality measures. Many (such as the ones listed above) expect data...

Note, SqlRender already supports DATEADD and some DATEDIFF calculations for non-day intervals. There is a [pull request](https://github.com/OHDSI/SqlRender/issues/345) to add support for Spark and Oracle - and it should be possible...

Based upon what I'm seeing while implementing the hypoglycemia measure, there are a few more small tweaks that may be needed or desirable. As context, I created cohorts using Atlas...

With the above detail table as a foundation, users are likely to want certain calculations. For example: * Time difference between two events (such as hours between qualifying event and...

Pull request #718 fixes these performance issues. Using the same data, queries 717, 815, and 1815 now take about 30 seconds each (on Spark) rather than minutes or hours.

@fdefalco and @chrisknoll , I refactored the queries as you instructed, and it is still good performance on Databricks. For our spark environment, we are using a Databricks "Small Warehouse"....