OHDSI Home | Forums | Wiki | Github

HowOften analysis is available for download and execution

General
#1

Dear OHDSI Community,

The HowOften analysis is now available for download from GitHub at the following location:

You can view the setup and execution instructions at the root of the repository in the README. The most important points is that you must be running R 4.2+, and have the specific R packages installed. The instructions contain information on setting up KeyRing with your connection details to your CDM, and instructions for executing the analysis.

Download the files by cloning the repository. Open the .R project file in R Studio and follow the instructions in the README.md.

We will be providing instructions on submitting results as we finalize the process, but you can begin executing these analyses as soon as you are ready.

Update Oct 19, 10:30AM: The George analysis json has been updated to fix an issue, if you have downloaded the repo prior to 10:30am on October 19th, please clone the repo again to get the updated file.

#2

A big thanks to everyone who attended the 2023 OHDSI Symposium and participated in the HowOften workshop. It was highly educational for me, and I enjoyed seeing the diverse environments that our community has to host and execute CDM analyses. Macs, Linux, Windows, Cloud hosted RStudio Server, Docker…many many things to consider!

For those of you who didn’t finish the run due to technical concerns, I hope that you will continue to push forward to get those pesky dependencies installed so that you can see the Strategus at work.

For those who have finished and would like to publish your results, please contact me at cknoll@ohdsi.org and I will provide instructions on how to package up your results.

I will be working on optimizing part of the analysis queries to work better on larger sets of data So, I hope that anyone who is still willing to participate may download a future update (hopefully 2 weeks) and re-run the analysis and compare execution times. I would also like to provide a Shiny app that would let you point to the results and view your results in the comfort of your own environment.

Thanks again for all your efforts.

-Chris

2 Likes
t