Some of the FRED data is monthly which is aggregated to quarterly. Currently the code just takes the end-of-period for each series, but some should be averages.

When new data is pulled, we should run some basic validation unit tests to make sure the new data looks correct. I am attaching a basic validation file that I created for the VFCI project that we can use as a starting point.

validate_data.R.zip

In the future, we will update the data. When this happens, the old data that is being replaced by the new data should be saved. Over time, we will accumulate several vintages that can allow us to "go back in time" if we need to, and also to do real-time analysis.

When changing a .rmd file, the GHA render-rmarkdown throughs an error.

Tried adding a line to load the packages "knitr" and "rmarkdown", but that did not solve the error.

For now, can simply render rmarkdown documents locally and push them.

See https://github.com/r-lib/actions/tree/v2/check-r-package

And get it to pass with the basic 1 function test package

Function that returns chart comparing each of the original bcadata and updated bcadata time series.

This will be the main function of the package.

It pulls all of the data from sources online.

• Function to pull FRED data (takes FRED key)
• Function to pull TFP data (if possible)
• Function to make the right transformations to the series
• Function that combines the above, returns new bcadata

I set up a Git Hub Action to test for code coverage in unit tests (i.e. are all functions executed during the testing process).

I think that to get the results to display nicely---including a badge on the ReadMe---we need to set up the repository on codecov, which is operated by Github. I tried going through and giving permissions, but I don't think I can do that for the VFCI organization repos.

Not pressing, just something to check out if we want.

Add unit tests with github action that fails if unit tests fail. As of now, I can only think of two, we can add more as we code or needs arise:

• bca data is replicated (#11)
• Data is validated (#9)

Add in the original, replication file BCA data from the authors.

We would like to replicate bca data more closely, hopefully exactly. A first try may be using ALFRED.