search

Big-Life-Lab / cchsflow

Variable transformation and harmonization for the Canadian Community Health Survey
https://big-life-lab.github.io/cchsflow/
Other
11 stars 7 forks source link

Derived variables + RData files for two CCHS cycles #8

Closed wyusuf068 closed 4 years ago

wyusuf068 commented 4 years ago

Completion of derived functions R file + vignette. Data folder created for 2003 and 2010 CCHS cycles as .RData files subsetted to 200 observations.

DougManuel commented 4 years ago

I've made edits to usingcchsflow.Rmd. These really aren't about derived variables (just reviewing the background leading to derived variables.

DougManuel commented 4 years ago

'DT' isn't working on my computer. The tables are all empty. I'd suggest using DT only when we need to. In usingcchsflow many of those datatables are small with few columns. They display pretty well without DT.

DougManuel commented 4 years ago

@wyusuf068 Note that I put a third option for harmonizing age. This is just to placemark our most common type of transformation.

DougManuel commented 4 years ago

I modified the comment about accessing cchs data at the very end of usingcchsdata. The whole example works well with the mock data that you created. But let's add an example using the 200 respondent sample at the end. (Doesn't need to be done for this PR).

DougManuel commented 4 years ago

I’ve tried moving the example of BMI over to 'usingcchsflow.Rmd'. What do you think?

It occurred to me that the user will harmonize BMI and other derived variables just like they do for any other variable. The only additional consideration for them is they need the underlying variables in their dataset (usually).

The Rmd's for variables.csv, variableDetails.csv and dervivedVariables are really reference guides (rather than a true vignette) and instructions about how to create new transformations.

DougManuel commented 4 years ago

I've made a few more edits to usingcchsflow.Rmd.

DougManuel commented 4 years ago

The PR looks good. Suggest merging and making any further modifications in a future PR.