From c2ad73743af0b6d70a0aa931c408eec139c4309a Mon Sep 17 00:00:00 2001
From: Michael Duong <michaelduong22@gmail.com>
Date: Thu, 8 Aug 2024 23:15:09 +0100
Subject: [PATCH] remove costing columns, photo supply and main-heating-control

---
 .../src/pipeline/configs/settings.yaml        | 10 ++--
 modules/ml-pipeline/src/pipeline/dvc.lock     | 53 +++++++++----------
 2 files changed, 29 insertions(+), 34 deletions(-)

diff --git a/modules/ml-pipeline/src/pipeline/configs/settings.yaml b/modules/ml-pipeline/src/pipeline/configs/settings.yaml
index aeb5907..f7b74be 100644
--- a/modules/ml-pipeline/src/pipeline/configs/settings.yaml
+++ b/modules/ml-pipeline/src/pipeline/configs/settings.yaml
@@ -39,9 +39,9 @@ default:
       drop_columns: ["hot_water_kwh"]
       retain_features: [
         'uprn', 
-        'heating-cost-current',
+        # 'heating-cost-current',
         'co2-emissions-current',
-        'hot-water-cost-current',
+        # 'hot-water-cost-current',
         'total-floor-area',
         'secondheat-description',
         'floor-description',
@@ -70,7 +70,7 @@ default:
         'number-open-fireplaces',
         'number-heated-rooms',
         'windows-description',
-        'photo-supply',
+        # 'photo-supply',
         'heat-loss-corridor',
         'flat-top-storey',
         'unheated-corridor-length',
@@ -84,9 +84,9 @@ default:
         'transaction-type',
         'floor-energy-eff',
         'lodgement-month',
-        'lighting-cost-current',
+        # 'lighting-cost-current',
         'glazed-area',
-        'main-heating-controls',
+        # 'main-heating-controls',
         'estimate_annual_kwh',
         ]
 
diff --git a/modules/ml-pipeline/src/pipeline/dvc.lock b/modules/ml-pipeline/src/pipeline/dvc.lock
index 8fd1507..1e29fb4 100644
--- a/modules/ml-pipeline/src/pipeline/dvc.lock
+++ b/modules/ml-pipeline/src/pipeline/dvc.lock
@@ -24,9 +24,7 @@ stages:
         - hot_water_kwh
         default.feature_processor.feature_processor_config.retain_features:
         - uprn
-        - heating-cost-current
         - co2-emissions-current
-        - hot-water-cost-current
         - total-floor-area
         - secondheat-description
         - floor-description
@@ -55,7 +53,6 @@ stages:
         - number-open-fireplaces
         - number-heated-rooms
         - windows-description
-        - photo-supply
         - heat-loss-corridor
         - flat-top-storey
         - unheated-corridor-length
@@ -69,9 +66,7 @@ stages:
         - transaction-type
         - floor-energy-eff
         - lodgement-month
-        - lighting-cost-current
         - glazed-area
-        - main-heating-controls
         - estimate_annual_kwh
         default.feature_processor.feature_processor_config.subsample_amount:
         default.feature_processor.feature_processor_config.subsample_seed: 0
@@ -87,8 +82,8 @@ stages:
     outs:
     - path: data/prepared_data/
       hash: md5
-      md5: c0cb87bd677d46f4c1a608c6d6ee6110.dir
-      size: 11062844
+      md5: a6241dcbb3fe1d3b39d1a300ea64dfc9.dir
+      size: 9606500
       nfiles: 2
   build_model:
     cmd: python 2_build_model.py
@@ -99,8 +94,8 @@ stages:
       size: 4820
     - path: data/prepared_data
       hash: md5
-      md5: c0cb87bd677d46f4c1a608c6d6ee6110.dir
-      size: 11062844
+      md5: a6241dcbb3fe1d3b39d1a300ea64dfc9.dir
+      size: 9606500
       nfiles: 2
     params:
       configs/build_model.yaml:
@@ -132,18 +127,18 @@ stages:
     outs:
     - path: data/fit_predictions/
       hash: md5
-      md5: 5c694b2ec23baca2e7242d3802ba09fe.dir
-      size: 1546012
+      md5: 5e07647b4dd0145a6d52d6ef729a3bde.dir
+      size: 1545562
       nfiles: 1
     - path: data/model/
       hash: md5
-      md5: 314c5043b1a421e847af8d3126fba788.dir
-      size: 278676507
-      nfiles: 37
+      md5: ce14e6f1e69c5513a04403eb00e0db0a.dir
+      size: 99464470
+      nfiles: 35
     - path: metrics/fit_metrics.json
       hash: md5
-      md5: 2e5ab3102b145f5c909232e66210a261
-      size: 222
+      md5: 425c0c6c13742d2d21051bf7ceb90127
+      size: 218
   generate_predictions:
     cmd: python 3_generate_predictions.py
     deps:
@@ -153,13 +148,13 @@ stages:
       size: 2464
     - path: data/model
       hash: md5
-      md5: 314c5043b1a421e847af8d3126fba788.dir
-      size: 278676507
-      nfiles: 37
+      md5: ce14e6f1e69c5513a04403eb00e0db0a.dir
+      size: 99464470
+      nfiles: 35
     - path: data/prepared_data
       hash: md5
-      md5: c0cb87bd677d46f4c1a608c6d6ee6110.dir
-      size: 11062844
+      md5: a6241dcbb3fe1d3b39d1a300ea64dfc9.dir
+      size: 9606500
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -171,8 +166,8 @@ stages:
     outs:
     - path: data/predictions/
       hash: md5
-      md5: 10b0612c10dfaa78e08ccc673c6f984e.dir
-      size: 163560
+      md5: ddaa04115c5dd4299974048080d762f5.dir
+      size: 163540
       nfiles: 1
   generate_metrics:
     cmd: python 4_generate_metrics.py
@@ -183,13 +178,13 @@ stages:
       size: 3484
     - path: data/predictions
       hash: md5
-      md5: 10b0612c10dfaa78e08ccc673c6f984e.dir
-      size: 163560
+      md5: ddaa04115c5dd4299974048080d762f5.dir
+      size: 163540
       nfiles: 1
     - path: data/prepared_data
       hash: md5
-      md5: c0cb87bd677d46f4c1a608c6d6ee6110.dir
-      size: 11062844
+      md5: a6241dcbb3fe1d3b39d1a300ea64dfc9.dir
+      size: 9606500
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -199,8 +194,8 @@ stages:
     outs:
     - path: metrics/metrics.json
       hash: md5
-      md5: 22794cfc31f6ffd98eb1db4852677b4f
-      size: 220
+      md5: 0beb72a28af4af37a619181b14c2e311
+      size: 218
   generate_scenerio_metrics:
     cmd: python 5_generate_scenarios.py
     deps: