Merge branch '42-canton-id-1' into 'develop'

impute canton id directly using shapefile instead of based on imputed municipalities

See merge request !86

data/spatial/cantons.py

-import pandas as pd
+import geopandas as gpd
 
 
 def configure(context):
     context.config("data_path")
 
+
 def execute(context):
     # Load data
     data_path = context.config("data_path")
 
-    df_cantons = pd.read_excel("%s/spatial_structure_2018.xlsx" % data_path,
-                               names=["municipality_id", "canton_id"],
-                               usecols=[0, 2],
-                               skiprows=6,
-                               nrows=2229,
-                               )
+    df = gpd.read_file("%s/municipality_borders/gd-b-00.03-875-gg18/ggg_2018-LV95/shp/g1k18.shp" % data_path,
+                       encoding="latin1"
+                       ).to_crs("EPSG:2056")
+
+    df = df.rename({"KTNR": "canton_id", "KTNAME": "canton_name"}, axis=1)
+    df = df[["canton_id", "canton_name", "geometry"]]
 
-    return df_cantons
+    return df
 
-def impute(df_cantons, df):
-    assert("municipality_id" in df.columns)
-    return pd.merge(df, df_cantons, on = "municipality_id", how = "left")
 
 SP_REGION_1 = [25, 12, 13, 1, 2, 14, 9]
 SP_REGION_2 = [21, 26, 15, 16, 22, 11, 24, 3, 6, 7]
 SP_REGION_3 = [17, 19, 10, 23, 20, 5, 18, 4, 8]
 
+
 def impute_sp_region(df):
-    assert("canton_id" in df.columns)
-    assert("sp_region" not in df.columns)
+    assert ("canton_id" in df.columns)
+    assert ("sp_region" not in df.columns)
 
     df["sp_region"] = 0
     df.loc[df["canton_id"].isin(SP_REGION_1), "sp_region"] = 1
@@ -39,5 +38,5 @@ def impute_sp_region(df):
     # Especially, we need a consistent spatial system. It probably would make
     # more sense to impute the SP region in another way
 
-    #assert(not np.any(df["sp_region"] == 0))
+    # assert(not np.any(df["sp_region"] == 0))
     return df

data/statpop/statpop.py

@@ -90,27 +90,39 @@ def execute(context):
     df_zones = context.stage("data.spatial.zones")
     df_municipality_types = context.stage("data.spatial.municipality_types")
     df_quarters = context.stage("data.spatial.quarters")
+    df_cantons = context.stage("data.spatial.cantons")
 
     df_spatial = pd.DataFrame(df[["person_id", "home_x", "home_y"]])
     df_spatial = data.spatial.utils.to_gpd(context, df_spatial, "home_x", "home_y")
 
+    # Impute municipalities
     df_spatial = data.spatial.utils.impute(context, df_spatial, df_municipalities, "person_id", "municipality_id")[[
         "person_id", "municipality_id", "geometry"
     ]]
+    df_spatial["municipality_id"] = df_spatial["municipality_id"].astype(np.int)
+
+    # Impute quarters
+    df_spatial = (data.spatial.utils.impute(context, df_spatial, df_quarters, "person_id", "quarter_id",
+                                            fix_by_distance=False)[
+        ["person_id", "municipality_id", "quarter_id", "geometry"]]
+    )
 
-    df_spatial = \
-        data.spatial.utils.impute(context, df_spatial, df_quarters, "person_id", "quarter_id", fix_by_distance=False)[[
-            "person_id", "municipality_id", "quarter_id", "geometry"
-        ]]
+    # Impute cantons
+    df_spatial = data.spatial.utils.impute(context, df_spatial, df_cantons, "person_id", "canton_id")[[
+        "person_id", "municipality_id", "quarter_id", "canton_id", "geometry"
+    ]]
 
+    # Impute municipality types
     df_spatial = data.spatial.municipality_types.impute(df_spatial, df_municipality_types)
+
+    # Impute zones
     df_spatial = data.spatial.zones.impute(df_spatial, df_zones)
 
     assert (len(df) == len(df_spatial))
 
     del df["municipality_id"]
     df = pd.merge(
-        df, df_spatial[["person_id", "zone_id", "municipality_type", "municipality_id", "quarter_id"]],
+        df, df_spatial[["person_id", "zone_id", "municipality_type", "municipality_id", "quarter_id", "canton_id"]],
         on="person_id"
     )
 
@@ -119,9 +131,6 @@ def execute(context):
     df["home_quarter_id"] = df["quarter_id"]
 
     # Impute SP region
-    df["municipality_id"] = df["municipality_id"].astype(np.int)
-    df_cantons = context.stage("data.spatial.cantons")
-    df = data.spatial.cantons.impute(df_cantons, df)
     df = data.spatial.cantons.impute_sp_region(df)
 
     # Impute population density