TODO: Harmonize & put source datafiles from csv to gdx

The source datafiles are partly redundant or in .csv format, which is storage inefficient. These should be put into gdx format (as they are not updated regularly or item by item).

Additionally, we should check whether yield data or xmat data is not redundant across datasets.

@michidoubleu

Start by inventorying the repository's source datafiles and identifying which are CSV, GDX, yield, or xmat datasets. Compare the yield and xmat data for redundancy, then define the harmonized GDX set and verify that the redundant data has been removed without losing required inputs.

Written by the indexing model from the issue text.