DISES: Proximity / Treatment vs Control villages (Cambodia)

Overview

• This repository contains Jupyter notebooks and supporting code to identify treatment and control villages relative to Community Forests (CFs), Forested Areas (FAs) and Community Protected Areas (CPAs). Key workflows: nearest-village treatment selection, fuzzy matching CF-reported villages to georeferenced villages, and multiple control-selection approaches (buffers, FA-only, etc).

Notebooks (open in VS Code / Jupyter)

• Controls.ipynb — end-to-end control-village selection examples and several approaches (A/B/C). Uses calculate_distances, find_nearest and buffer workflows.
  • referenced symbols: calculate_distances, find_nearest
• Proximity_to_CFs.ipynb — raster filtering, forest-area shapefile creation, and initial nearest-village treatment selection. Implements filter_raster_by_percentage_and_area and raster → polygon extraction.
  • referenced symbols: [filter_raster_by_percentage_and_area](Proximity to CFs.ipynb), calculate_distances, find_nearest
• Villages_Matching.ipynb — fuzzy matching CF-reported village names to georeferenced village dataset (uses thefuzz / rapidfuzz), builds treatment set and falls back to nearest-village for non-matches.
  • referenced symbols: find_closest_phumcode
• Villages_Matching-NIS.ipynb — variant that joins NIS village table into the matching workflow.

Key scripts and resources

• Data processing helper (MPI postprocessing): [data/STATA DHS/stata_code/code/MPI/postprocess_summarize_mpi.py](data/STATA DHS/stata_code/code/MPI/postprocess_summarize_mpi.py)
• WHO/IGROWUP STATA macros and reference tables: [data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/README.md](data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/README.md) and license [data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/LICENSE](data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/LICENSE)
  • STATA macros of interest: [data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/igrowup_standard.ado](data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/igrowup_standard.ado), [data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/igrowup_restricted.ado](data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/igrowup_restricted.ado)
  • WHO 2007 reference: [data/STATA DHS/stata_code/code/MPI/ado/who2007_stata/who2007.ado](data/STATA DHS/stata_code/code/MPI/ado/who2007_stata/who2007.ado)

Data folders (examples)

• Root data directory used in notebooks: data/
• Example shapefiles referenced by notebooks:
  • Updated CFs: Camboda CF updated data - Feb 5, 2025/CF_KH_Updated_2.5.2025_EPSG32648.shp & data/Camboda CF updated data - Feb 5, 2025-20250217T220026Z-001.zip (zipped in data/ original shapefile inside)
  • Villages: KHM-Villages/Villages.shp
  • CPA shapefile: Community_Protected_Areas_Cambodia/CPA_Shape_31_Aug_2022.shp

How to run (recommended)

1. Open the desired notebook in VS Code (Jupyter) and run cells interactively.
2. Ensure the notebooks' root_path variables point to the repository root (many cells use root_path = 'DISES/Proximity/data' or relative data/ paths). Adjust if running from a different working directory.
3. Required Python packages (install via pip or conda):
   • geopandas, pandas, shapely, rasterio, numpy, scipy, matplotlib, contextily, rapidfuzz, thefuzz
4. For heavy spatial ops (buffers, sjoin), use a machine with sufficient memory; consider simplifying geometries before buffering (notebooks show .simplify usage).

Notes & pointers

• Fuzzy matching threshold is configurable (see variable threshold in Villages_Matching.ipynb).
• The notebooks create GeoDataFrames that are exported as shapefiles under outputs paths (check outputs_path or hardcoded to_file calls).
• STATA macros and license files are kept under data/STATA DHS/stata_code/code/MPI/ado/igrowup_stata/ — consult the README and LICENSE if you reuse those macros.
• If you need to find where a function is defined in the repo, search the notebook file; e.g. calculate_distances is defined inline in the notebooks.