Jason-Perez27/SSA-Pivot-Detect

Repository for detecting activity status in center pivots across Sub-Saharan Africa. Leveraging ML to understand agricultural practices # SSA-Pivot-Detect Repository for detecting activity status in center pivots across Sub-Saharan Africa. Leveraging ML to understand agricultural practices Authors: Jason Perez, Anna Boser, Kelly Caylor ## Setup The code in this directory is written in Python, and we use anaconda and a requirements.txt to manage our packages. Download anaconda and run the following commands in your terminal to set up your environment: ```{bash} conda create -n cp_pipeline python=3.9 -y conda activate cp_pipeline pip install -r requirements.txt ``` ## Shapefile containing Center Pivots across the World You can obtain the shp file containing the world's center pivots here. To extract the Center Pivots identified in 2021, download all files in this folder and run the following code in the terminal: ```{bash} cd ~/Downloads zip -s 0 World_CPIS_2021.zip --out World_CPIS_2021_together.zip unzip World_CPIS_2021_together.zip ``` ## Inventory of the data folder Because many of the data used in this project are too large to upload to GitHub, the data folder is in the .gitignore file and therefore none of its contents are uploaded. However, here we include a description of all datasets and their location within the data folder. These data can also be found in this google drive folder. - World_CPIS_2021 - The original World CPIS data from this folder - A modified version of this dataset with IDs, created in `code/cp_data_subset/1_id_cp.py` - Only the CPIS to be used for training and testing (whose IDs appear in `data/cp_ids.txt`). Created in `code/cp_data_subset/3_data_subset.py` - 1_script_data: Includes files contained from scripts within `1_cp_data_subset` - id_cp_shapefile.shp: Same file as `World_CPIS_2021` but IDs assigned to each center pivot. Used in `code/cp_data_subset/1_id_cp.py` - cp_ids.txt: A list of CP IDs to be used in the training and testing of the algorithm. Determined in `code/cp_data_subset/2_id_filter_cp.py`. - map.geojson: GeoJSON file containing a geometric shape …