From c6f869e95fad395ff7c7fe61fff2c853b2b4bcdf Mon Sep 17 00:00:00 2001
From: Andras Schmelczer <andras@schmelczer.dev>
Date: Mon, 2 Feb 2026 22:28:51 +0000
Subject: [PATCH] Add tax bands and update journey times

---
 Taskfile.data.yml | 43 ++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 38 insertions(+), 5 deletions(-)

diff --git a/Taskfile.data.yml b/Taskfile.data.yml
index 5907545..4e6bd4c 100644
--- a/Taskfile.data.yml
+++ b/Taskfile.data.yml
@@ -11,7 +11,7 @@ vars:
   EPC_PP_OUTPUT: "{{.DATA_DIR}}/epc_pp.parquet"
   WIDE_OUTPUT: "{{.DATA_DIR}}/wide.parquet"
   EPC: "{{.DATA_DIR}}/certificates.csv"
-  JOURNEY_TIMES: "./data_sources/processed/journey_times_bank_checkpoint.parquet"
+  JOURNEY_TIMES: "{{.DATA_DIR}}/journey_times.parquet"
   ETHNICITY_OUTPUT: "{{.DATA_DIR}}/ethnicity_by_la.parquet"
   CRIME_DIR: "{{.DATA_DIR}}/crime"
   CRIME_OUTPUT: "{{.DATA_DIR}}/crime_by_lsoa.parquet"
@@ -20,6 +20,8 @@ vars:
   NAPTAN_OUTPUT: "{{.DATA_DIR}}/naptan.parquet"
   BROADBAND_OUTPUT: "{{.DATA_DIR}}/broadband.parquet"
   SCHOOL_PROXIMITY_OUTPUT: "{{.DATA_DIR}}/school_proximity.parquet"
+  COUNCIL_TAX_OUTPUT: "{{.DATA_DIR}}/council_tax.parquet"
+  COUNCIL_TAX_BANDS_OUTPUT: "{{.DATA_DIR}}/council_tax_bands.parquet"
 
 tasks:
   prompt:epc:
@@ -107,6 +109,20 @@ tasks:
     cmds:
       - uv run python -m pipeline.download.broadband --output {{.BROADBAND_OUTPUT}}
 
+  download:council-tax:
+    desc: Download council tax rates by local authority (GOV.UK Table 9)
+    status:
+      - test -f {{.COUNCIL_TAX_OUTPUT}}
+    cmds:
+      - uv run python -m pipeline.download.council_tax --output {{.COUNCIL_TAX_OUTPUT}}
+
+  download:council-tax-bands:
+    desc: Scrape individual property council tax bands from VOA
+    status:
+      - test -f {{.COUNCIL_TAX_BANDS_OUTPUT}}
+    cmds:
+      - uv run python -m pipeline.download.council_tax_bands --postcodes {{.WIDE_OUTPUT}} --output {{.COUNCIL_TAX_BANDS_OUTPUT}}
+
   download:noise:
     desc: Download Defra noise data (road, rail, airport) sampled at postcode centroids
     deps:
@@ -164,13 +180,15 @@ tasks:
       - uv run python -m pipeline.transform.school_proximity --ofsted {{.OFSTED_OUTPUT}} --arcgis {{.ARCGIS_OUTPUT}} --output {{.SCHOOL_PROXIMITY_OUTPUT}}
 
   download:journey-times:
-    desc: Fetch TfL journey times for all postcodes
+    desc: "Fetch TfL journey times: task download:journey-times -- <destination>"
     deps:
       - download:arcgis
+    requires:
+      vars: [CLI_ARGS]
     status:
-      - test -f {{.JOURNEY_TIMES}}
+      - test -f {{.DATA_DIR}}/journey_times_{{.CLI_ARGS}}.parquet
     cmds:
-      - uv run python -m pipeline.journey_times
+      - uv run python -m pipeline.journey_times --destination {{.CLI_ARGS}} --output-dir {{.DATA_DIR}} --postcodes {{.ARCGIS_OUTPUT}}
 
   prepare:
     desc: Build wide property dataframe with all joins
@@ -181,6 +199,7 @@ tasks:
       - download:ethnicity
       - download:broadband
       - download:noise
+      - download:council-tax
       - transform:crime
       - transform:poi-proximity
       - transform:school-proximity
@@ -188,4 +207,18 @@ tasks:
     status:
       - test -f {{.WIDE_OUTPUT}}
     cmds:
-      - uv run python -m pipeline.transform.merge --epc-pp {{.EPC_PP_OUTPUT}} --arcgis {{.ARCGIS_OUTPUT}} --iod {{.IOD_OUTPUT}} --poi-proximity {{.POI_PROXIMITY_OUTPUT}} --journey-times {{.JOURNEY_TIMES}} --ethnicity {{.ETHNICITY_OUTPUT}} --crime {{.CRIME_OUTPUT}} --noise {{.NOISE_OUTPUT}} --school-proximity {{.SCHOOL_PROXIMITY_OUTPUT}} --broadband {{.BROADBAND_OUTPUT}} --output {{.WIDE_OUTPUT}}
+      - >-
+        uv run python -m pipeline.transform.merge
+        --epc-pp {{.EPC_PP_OUTPUT}}
+        --arcgis {{.ARCGIS_OUTPUT}}
+        --iod {{.IOD_OUTPUT}}
+        --poi-proximity {{.POI_PROXIMITY_OUTPUT}}
+        --journey-times {{.JOURNEY_TIMES}}
+        --ethnicity {{.ETHNICITY_OUTPUT}}
+        --crime {{.CRIME_OUTPUT}}
+        --noise {{.NOISE_OUTPUT}}
+        --school-proximity {{.SCHOOL_PROXIMITY_OUTPUT}}
+        --broadband {{.BROADBAND_OUTPUT}}
+        --council-tax {{.COUNCIL_TAX_OUTPUT}}
+        --council-tax-bands {{.COUNCIL_TAX_BANDS_OUTPUT}}
+        --output {{.WIDE_OUTPUT}}