WIP auto convert old cons data to new cons data

hub/management/commands/create_new_con_data_from_old.py

@@ -0,0 +1,85 @@
+from django.core.management.base import BaseCommand
+
+import pandas as pd
+from mysoc_dataset import get_dataset_df
+
+from hub.models import Area, AreaData, AreaType, DataSet, DataType
+from utils.constituency_mapping import convert_data_geographies
+
+
+class Command(BaseCommand):
+    help = "Create new constituency data from old constituency data"
+
+    new_con_at = AreaType.objects.get(code="WMC23")
+
+    def add_arguments(self, parser):
+        parser.add_argument(
+            "-q", "--quiet", action="store_true", help="Silence progress bars."
+        )
+
+    def fetch_parl25_gss_map(self):
+        df = get_dataset_df(
+            repo_name="2025-constituencies",
+            package_name="parliament_con_2025",
+            version_name="latest",
+            file_name="parl_constituencies_2025.csv",
+        )
+        self.parl25_gss_map = df.set_index("short_code").gss_code.to_dict()
+
+    def apply_parl25_gss_to_df(self, df):
+        df["PARL25"] = df["PARL25"].apply(
+            lambda name: self.parl25_gss_map.get(name, None)
+        )
+        return df
+
+    def get_df_from_dataset(self, ds):
+        if ds.table == "areadata":
+            data = AreaData.objects.filter(
+                data_type__data_set=ds, area__area_type__code="WMC"
+            )
+
+        data_list = []
+        for d in data:
+            data_list.append([d.area.gss, d.value()])
+
+        df = pd.DataFrame(data_list)
+        df.columns = ["PARL10", "value"]
+
+        return df
+
+    def create_data_for_new_con(self, ds, df):
+        dt = DataType.objects.get(data_set=ds)
+        for _, row in df.iterrows():
+            a = Area.objects.get(gss=row["PARL25"], area_type__code="WMC23")
+            AreaData.objects.update_or_create(
+                area=a,
+                data_type=dt,
+                defaults={
+                    "float": row["value"],
+                },
+            )
+        ds.areas_available.add(self.new_con_at)
+
+    def process_datasets(self):
+        sets = DataSet.objects.filter(
+            unit_distribution="people_in_area",
+            category__in=["place", "opinion"],
+            is_range=False,
+        )
+
+        for ds in sets:
+            print(ds.label)
+            df = self.get_df_from_dataset(ds)
+            new_df = convert_data_geographies(
+                df=df,
+                input_geography="PARL10",
+                output_geography="PARL25",
+                input_values_type=ds.unit_type,
+            )
+            new_df = self.apply_parl25_gss_to_df(new_df)
+            self.create_data_for_new_con(ds, new_df)
+
+    def handle(self, quiet=False, *args, **options):
+        self._quiet = quiet
+        self.fetch_parl25_gss_map()
+        self.process_datasets()