Restructure project folder

[matdehaven]

The coding folder has become somewhat unwieldy.

Restructure and organize the project folder.

• [x] Main.r only runs necessary scripts to make paper-figure charts
• [x] Separate BCA replication script
• [ ] GHA that runs reports
• [ ] Reports converted to Quarto (compile to html?)
• [ ] Set up a website in docs/ folder
• [x] Subfolder as an R package, instead of documentation in main folder?

[matdehaven]

"vici-bc" folder changed to "run-vfci-bc-analysis" folder and creates the data files needed for the paper figures and tables.

Working next on the mean-vol calculations, adding a function to estimate the heteroskedasticity from a VAR.

[matdehaven]

The main script is now done, producing all of the data files needed to make the figures in the paper. The code is much better structured and easier to read. There is now a function to calculate the mean-vol estimates from a VAR.

[matdehaven]

Following items still need to be completed:

• [x] Make paper figures from the paper-figures data
• [x] Store paper-figures in the repository and then copy to Overleaf
• [x] GHA that runs the BCA replication and related tests leave tests where they are for now
• [x] Add tests for paper claims

[matdehaven]

Main.r script is done and recreates the figures and tables used in the current draft of the paper.

There are still loose ends (namely, the reports), but mostly the project structure is now in a state that I am happy with.

[matdehaven]

The overleaf submodule seems to no longer be working. Need to relink it and then update the links to all of the tables and figures.

[matdehaven]

Fixed the submodule. Somehow the submodule was pointing to a specific commit/branch and to the Overleaf "master" branch. I was able to change that in VS Code and then commit and push to Overleaf.

[matdehaven]

Following Shapiro and Gentzkow RA Manual use "autofilling values" by defining latex commands for any specific numbers referenced in the text (correlations, etc.) with R code. Then reference those values in Latex by using the command.

I am calling these "inline-values" and they are saved under the figures/ folder.

These are obvious candidates for having a unit test for each.

[matdehaven]

Setup a testing script that tests the inline values against what we expect them to be.

Currently inline values does not cover all of the explicit numbers referenced in the paper, but it is enough to have the structure settled for the next draft.

[matdehaven]

All the remains is to tackle the "reports" folder. This needs to be reorganized for the new project structure, where the idea is that each "report" file is a separate, stand-alone script that relies minimally on the project files itself (ie, reads raw data, not final analysis data). The idea being that these scripts are for testing out new ideas and shouldn't be maintained along with the main.r code.

For now, I am moving this task to low priority and will revisit this after finishing the current state report. That should give me a sense of the structure I would like to use moving forward.

Still need to decide whether to include a website or not for the project.