From 1ad96ed3145e2165c8a663697b18dbec4cd1bddb Mon Sep 17 00:00:00 2001
From: Michael Duong <michaelduong22@gmail.com>
Date: Fri, 9 Aug 2024 10:54:22 +0100
Subject: [PATCH] remove the same feature: cost, photo and mainheat

---
 .../src/pipeline/configs/settings.yaml        | 10 ++--
 modules/ml-pipeline/src/pipeline/dvc.lock     | 51 +++++++++----------
 2 files changed, 28 insertions(+), 33 deletions(-)

diff --git a/modules/ml-pipeline/src/pipeline/configs/settings.yaml b/modules/ml-pipeline/src/pipeline/configs/settings.yaml
index e24abf5..f623210 100644
--- a/modules/ml-pipeline/src/pipeline/configs/settings.yaml
+++ b/modules/ml-pipeline/src/pipeline/configs/settings.yaml
@@ -33,9 +33,9 @@ default:
       drop_columns: ["heating_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',
@@ -64,7 +64,7 @@ default:
         'number-open-fireplaces',
         'number-heated-rooms',
         'windows-description',
-        'photo-supply',
+        # 'photo-supply',
         'heat-loss-corridor',
         'flat-top-storey',
         'unheated-corridor-length',
@@ -78,9 +78,9 @@ 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 ff64aca..1061d02 100644
--- a/modules/ml-pipeline/src/pipeline/dvc.lock
+++ b/modules/ml-pipeline/src/pipeline/dvc.lock
@@ -24,9 +24,7 @@ stages:
         - heating_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: 19abfc15e24cd8a869a0f3f087e09584.dir
-      size: 13396685
+      md5: c45c73e2e25a5c9697a788cfa04f232d.dir
+      size: 11682246
       nfiles: 2
   build_model:
     cmd: python 2_build_model.py
@@ -99,8 +94,8 @@ stages:
       size: 4820
     - path: data/prepared_data
       hash: md5
-      md5: 19abfc15e24cd8a869a0f3f087e09584.dir
-      size: 13396685
+      md5: c45c73e2e25a5c9697a788cfa04f232d.dir
+      size: 11682246
       nfiles: 2
     params:
       configs/build_model.yaml:
@@ -132,17 +127,17 @@ stages:
     outs:
     - path: data/fit_predictions/
       hash: md5
-      md5: 0e2f1131b24eafa09ccbab91a36cac6c.dir
-      size: 1787805
+      md5: 6abffc8f19e3bb14345f0504a96fd214.dir
+      size: 1788386
       nfiles: 1
     - path: data/model/
       hash: md5
-      md5: 77c0900cda64c1beaef6782d04fd712e.dir
-      size: 518798187
-      nfiles: 36
+      md5: aee2886545c62efbf26d49f32bd1f328.dir
+      size: 79940408
+      nfiles: 35
     - path: metrics/fit_metrics.json
       hash: md5
-      md5: 0e03bb46cd03e9521095bf141ee92ed7
+      md5: 14e5b4019f6e5cf49edf7945b71e6a66
       size: 220
   generate_predictions:
     cmd: python 3_generate_predictions.py
@@ -153,13 +148,13 @@ stages:
       size: 2464
     - path: data/model
       hash: md5
-      md5: 77c0900cda64c1beaef6782d04fd712e.dir
-      size: 518798187
-      nfiles: 36
+      md5: aee2886545c62efbf26d49f32bd1f328.dir
+      size: 79940408
+      nfiles: 35
     - path: data/prepared_data
       hash: md5
-      md5: 19abfc15e24cd8a869a0f3f087e09584.dir
-      size: 13396685
+      md5: c45c73e2e25a5c9697a788cfa04f232d.dir
+      size: 11682246
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -171,8 +166,8 @@ stages:
     outs:
     - path: data/predictions/
       hash: md5
-      md5: efc7b9dd9d40e42ad93e3e5acbeac92d.dir
-      size: 192502
+      md5: efe40990a6092494363daa3284a22878.dir
+      size: 192442
       nfiles: 1
   generate_metrics:
     cmd: python 4_generate_metrics.py
@@ -183,13 +178,13 @@ stages:
       size: 3484
     - path: data/predictions
       hash: md5
-      md5: efc7b9dd9d40e42ad93e3e5acbeac92d.dir
-      size: 192502
+      md5: efe40990a6092494363daa3284a22878.dir
+      size: 192442
       nfiles: 1
     - path: data/prepared_data
       hash: md5
-      md5: 19abfc15e24cd8a869a0f3f087e09584.dir
-      size: 13396685
+      md5: c45c73e2e25a5c9697a788cfa04f232d.dir
+      size: 11682246
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -199,8 +194,8 @@ stages:
     outs:
     - path: metrics/metrics.json
       hash: md5
-      md5: a8b7699dd2b171da5fead1fb04d954aa
-      size: 220
+      md5: c6f913d497eb2f98e801c9e030bd96e9
+      size: 222
   generate_scenerio_metrics:
     cmd: python 5_generate_scenarios.py
     deps: