As we are expecting new indicators to be added, and the VI to continually adapt to new information, I tried to automate the generation of metadata files as much as possible:
Created a metadata folder
Moved old .md metadata file to new metadata folder
Created an R Script that generates metadata templates for the overall VI (by MSOA), and the two bespoke VI's (food and hardship). For the overall VI, these templates guess the domain of a given indicator based off of its position (i.e., row index) and name in the VI.
The source and date of the indicators (and domains for the bespoke VI's) needs adding after the generation of the templates. The metadata-match-sources.R file does this for existing indicators, but any new indicators will need manually adding. See the README.md instructions in the /metadata folder for more info.
For this first iteration, I have manually retrieved all the sources and dates. There are several discrepencies between the Google Doc listing the indicators and the actual indicators that feature in the different VI's (e.g., the Google Doc says the hardshup fund VI was just for England + Wales, which is no longe true). To retrieve the relevant sources and dates, the Google Doc and code for the VI's were searched for links and references.
As we are expecting new indicators to be added, and the VI to continually adapt to new information, I tried to automate the generation of metadata files as much as possible:
metadata-match-sources.Rfile does this for existing indicators, but any new indicators will need manually adding. See theREADME.mdinstructions in the/metadatafolder for more info.