We now have an evil mix of pickle files, JSON files, and SQLite databases. It is very, very dumb. The original justification was that loading 500.000 locations codes (during regionalized LCA) was much faster in one file than a big database query, and JSON allowed configurations to be edited without having to know how to use a relational database. These reasons are inadequate.
We now have an evil mix of pickle files, JSON files, and SQLite databases. It is very, very dumb. The original justification was that loading 500.000 locations codes (during regionalized LCA) was much faster in one file than a big database query, and JSON allowed configurations to be edited without having to know how to use a relational database. These reasons are inadequate.
Specific tasks:
bw2data
to SQLite databasesNice-to-have: