search

opensafely / school-age-children-and-covid

Repo for the project that examines the association between school age children in the house and covid-19 outcomes
3 stars 0 forks source link

add worms variable #21

Closed hjforbes closed 4 years ago

hjforbes commented 4 years ago

Hi @CarolineMorton and @alexwalkercebm

So I've added the worms variable. However, we think we might need to run this analysis on a tweaked study pop, which will require a new study definition. It will be almost identical to the existing analysis, but starting feb 1st 2019.

Wondering what the best way to do this is, without introducing errors! A separate folder perhaps? Any thoughts would be ace

FYI @laurietomlinson

CarolineMorton commented 4 years ago

I think if you want a new study for comparison, copy the study_definition.py file and name it something like study_definition_[addition_name].py. This output a csv file called input_[addition_name].csv which you can refer to in your stata. I will let you do this before I review as it will be easier that way.

See NSAIDs repo for example.

hjforbes commented 4 years ago

@CarolineMorton Okay, thanks. I'll need to repeat do files 1-8 as well. Can I keep all the worm-related files (study def and do files) within a sub-folder in the analysis folder? I'm slightly concerned that I'm creating a lot of do files to check/keep track of/ensure are consistent, and it might get messy. But maybe there's no better way of doing this?

CarolineMorton commented 4 years ago

The study_definition.py files need to remain in the analysis folder and so does the model.do. All other files can live in sub folders (however you decide) and pathed to in stata.

hjforbes commented 4 years ago

Apologies for committing the output files again....