UPDATE 31/10/2024 AUTHOR cmbrennan002

I've updated the files to tidy up work done previously and added a README. Changes made since the first round of work include

• Reconciling the products back to original set used (mostly dropping products that were missing NPM information)
• Tidying up the scripts to one script that creates the model input and another which runs the model
• Applying some initial specific gravity conversions so that the products measured in litres are treated appropriately for a SWA

To complete this review, please:

• Review the README (which should explain the steps in the files below)
• Review logic in <ahl_targets/analysis/change_checks_Oct24/compare_new_to_old_file.py>
• Review and check <create_new_datafile.py> runs
• Further investigate the model outputs we're getting

Investigations still required are:

• Checking the specific gravity - should we apply it at rst_extended level, and is there an original file we can use rather than my quickly googled values?
• The volume adjustment (with specific gravity) leads to a change in the input SWA, but the same new SWA (and kcal reduction pp) after the model is run - need to get to the bottom of this.
  - I wonder if this is because the coefficients haven't been updated yet?
• There are ~19,000 cooking oils with ED>900 - need to check if this makes sense or they should be dropped

Then general sense checks and analysis of how the SWA by store and market change - I've begun this in <ahl_targets/analysis/change_checks_Oct24/new_swa_analysis.py>

Description

Edit:

Please see pipeline/model_data_diets.py for the script to generate the input data inline with diets specs.

Hey both, @cmbrennan002 I know you've already had a look but just creating this so you can also have a scan @izzyStewart.

All changes to review are in the file ahl_targets/pipeline/simulation_npm_v2.py. The big question is why the updated baseline is so much smaller than in diets. Let me know if you have any thoughts on why that's the case!

Fixes #98