- Created `environment_pytorch.yml` to define the conda environment for PyTorch with GPU support, including necessary packages like GDAL and SentinelHub.
- Added `requirements_harvest.txt` for Python package dependencies related to data harvesting, including PyTorch, Pandas, NumPy, and others.
- Renamed translation function from `t` to `tr_key` for clarity and consistency.
- Updated all instances of translation calls in `90_CI_report_with_kpis_agronomic_support.Rmd` and `91_CI_report_with_kpis_cane_supply.Rmd` to use `tr_key`.
- Introduced a new helper function `get_area_unit_label` to manage area unit preferences across the project.
- Modified area calculations in `91_CI_report_with_kpis_cane_supply.Rmd` to utilize area from analysis data instead of recalculating.
- Added area unit preference setting in `parameters_project.R` to allow for flexible reporting in either hectares or acres.
- Updated `MANUAL_PIPELINE_RUNNER.R` to include language parameter for report generation.
- Adjusted translations in the `translations.xlsx` file to reflect changes in the report structure.
- Introduced core and optional features for the code improvements merge.
- Documented detailed implementation checklist for core features including database migration, model updates, job changes, form modifications, shell script wrappers, and Python package files.
- Added optional enhancements such as country-based project organization, download scheduling, project search feature, and harvest prediction setup documentation.
- Included testing checklist and post-merge data recreation strategy.
Timon