This component contains the Python notebooks and scripts to process the code.
*Note that file paths will need to be updated to run any of these scripts*
Python scripts in `code` (all contain functions used by Jupyter notebooks):
- `ask_RQs.py` : Functions needed by the jupyter notebooks for analysis.
- `names.py` : Dictionary to format the variable names
- `read_gee_annual.py`: Functions needed to read csv files exported from GEE; see notebooks for implementation.
`visuals.py` : functions to create visuals.
Processing notebooks:
- Process annual GEE data.ipynb: processes the output CSV files from GEE and saves to a pandas pkl. Note that the raw csv files are note included here, but can be repoduced with the GEE script `export_annual_mod11.js`
Exploratory / illustrative notebooks:
- `Covariates marginal.ipynb`: creates scatter plots of the 'local' covariates (L, FC, and eta), shown with marginal distriubtions (Figure 4 in the manuscript).
- `Exploratory, boxplots.ipynb` : creates exploratory scatter plots
- `visuals project.ipynb`: creates visuals illustrating how the data was sampled and balanced for Research Question 1 (used in Figure 6).
Research question 1 notebooks include:
- `RQ1 Panels A and B.ipynb`
- `RQ1 Panel C.ipynb`
- `RQ1 Panel D.ipynb`
Research quation 2 notebooks include:
1. `RQ2 5% filter; results.ipynb`: assessess the sensitivity of warming to the spatial extent over which forest loss occurs (RQ 2 analysis; Figure 8a in the manuscript).
2. `RQ2 2% filter; robustness.ipynb`, `RQ2 2% filter; no R-10km filter.ipynb` ... (all other notebooks starting with `RQ2`) : robustness checks for research question 2.
Research question 3: notebooks include:
1. `RQ3 nonlocal 2% filter.ipynb` : assesses the sensitivity of temperature increase to nonlocal forest loss (RQ 3 analysis; Figure 8B).
