combine.py

import pandas as pd
import geopandas as gpd


snotel_df = pd.read_csv('./processed/out.csv')
met_df = pd.read_csv('./processed/met_out.csv')
print('pandas loaded')


met_gdf = gpd.GeoDataFrame(met_df, geometry=gpd.points_from_xy(met_df.lat, met_df.lon))
snotel_gdf = gpd.GeoDataFrame(snotel_df, geometry=gpd.points_from_xy(snotel_df.lat, snotel_df.lon))
print('geopandas loaded')


shards = {k:d for k, d in met_gdf.groupby("date")}
print('shards loaded')
# print(shards['2000-01-01'])

def runpershard(d):
    datekey = d["Date"].values[0]
    print(datekey)
    return gpd.sjoin_nearest(d, shards[datekey], distance_col="distance_from_met")

df_n = snotel_gdf.groupby("Date").apply(runpershard)
print('done with merge')
df_n.to_csv("./processed/combined.csv")
print(df_n)

# OLD VERSION ================================================
# import pandas as pd
# import geopandas as gpd

# snotel_df = pd.read_csv('./processed/out.csv')
# met_df = pd.read_csv('./processed/met_out.csv')
# met_df = met_df.loc[met_df["date"]=="1990-01-01"]
# met_gdf = gpd.GeoDataFrame(met_df, geometry=gpd.points_from_xy(met_df.lat, met_df.lon))
# exp = met_gdf.explore(height=300,width=400)
# outfp = r"processed/explore.html"
# exp.save(outfp)

# snotel_gdf = gpd.GeoDataFrame(snotel_df, geometry=gpd.points_from_xy(snotel_df.lat, snotel_df.lon))
# print("loaded")
# # ---------------
# gdf = gpd.sjoin_nearest(snotel_gdf, met_gdf, how="inner", distance_col="dist_to_met")
# print(gdf)