build_csv*.py are used to build CSV files from netCDF4 files on Gates/Local machine.
smooth-result-build-npy.py read predictions and calculate R2 score for different models.
smooth-result.py read R2 npy file and plot box and line plots.
smooth-spatial.py plot spatial distribution of concurrent R2.
smooth-spatial-lag.py plot spatial distribution of lag R2.
model-train-smooth.py train models with args (pre_season, smooth, eof, lag (default0)).
include/ for reduced models.

real-smooth.py apply ML models with USGS and reanalysis dataset.

train_full.py train ML models with historical CMIP6 in a cross-validation setting.
train_full_all.py train ML models with historical CMIP6 and cross-validation with multiprocessing.
train_full_ssp.py train ML models with historical+ssp CMIP6 in a train-test setting.

Figures/ plot scripts.
dataResults aggregates data and calculate R2 scores.