modules/ml-pipeline/src/pipeline/4_generate_metrics.py

@@ -4,7 +4,9 @@ After the model is built, we can evaluate its performance
 """
 
 import os
+import yaml
 import pandas as pd
+from pathlib import Path
 from core.interface.InterfaceModels import MLModel
 from core.interface.InterfaceMetrics import MLMetrics
 from core.interface.InterfaceDataClient import DataClient

modules/ml-pipeline/src/pipeline/configs/feature_processor_logic.py

@@ -18,39 +18,30 @@ def remove_starting_columns(df):
     return df
 
 
-def keep_negative_heat_change(df):
-    df = df[df["heat_demand_change"] < 0]
+def remove_floor_height_ending(df):
+    # df.describe(percentiles=[0.005,0.99])['FLOOR_HEIGHT_ENDING']
+    # shows bottom 0.5 percentile is 1.665
+    # So keep anything above this
+    df = df[df["floor_height_ending"] > 1.665].reset_index(drop=True)
+    print("we in here")
     return df
 
 
-def keep_negative_carbon_change(df):
-    df = df[df["carbon_change"] < 0]
+def remove_minimum_habitable_room_size(df):
+    # Need minimum of 6.5m per habitable room
+    df = df[
+        df["total_floor_area_ending"] / df["number_habitable_rooms"] > 6.5
+    ].reset_index(drop=True)
     return df
 
 
-# TODO: Move to ETL pipeline
-def remove_unreasonable_habitable_rooms(df):
-    """
-    Assumption is that proportion of floor area to habitable rooms should be at least 6.5m2
-    """
-    minimum_room_size_index = (
-        df["total_floor_area_ending"] / df["number_habitable_rooms"] >= 6.5
-    )
-    df = df[minimum_room_size_index]
+def keep_flats(df):
+    df = df[df["property_type"] == "Flat"]
     return df
 
 
-def remove_top_1_percent_heat_demand(df):
-    # threshold_value = df.describe(percentiles=[0.99])['HEAT_DEMAND_STARTING']['99%']
-    threshold_value = 860
-    df = df[df["heat_demand_starting"] < threshold_value]
-    return df
-
-
-def remove_top_1_percent_carbon(df):
-    # threshold_value = df.describe(percentiles=[0.99])['CARBON_STARTING']['99%']
-    threshold_value = 18
-    df = df[df["carbon_starting"] < threshold_value]
+def keep_non_zero_rdsap(df):
+    df = df[df["rdsap_change"] != 0]
     return df
 
 
@@ -63,11 +54,10 @@ def remove_top_1_percent_carbon(df):
 #     return df
 
 business_logic = {
-    "remove_unreasonable_habitable_rooms": remove_unreasonable_habitable_rooms,
-    "keep_negative_heat_change": keep_negative_heat_change,
-    "keep_negative_carbon_change": keep_negative_carbon_change,
-    "remove_top_1_percent_heat_demand": remove_top_1_percent_heat_demand,
-    "remove_top_1_percent_carbon": remove_top_1_percent_carbon,
+    # "keep_non_zero_rdsap": keep_non_zero_rdsap,
+    # "keep_flats": keep_flats,
+    # "remove_minimum_habitable_room_size": remove_minimum_habitable_room_size,
+    # "remove_floor_height_ending": remove_floor_height_ending
     # "remove_starting_columns": remove_starting_columns
     # "keep_ENDING_COLUMNS": keep_ending_columns
 }

modules/ml-pipeline/src/pipeline/configs/post_prediction_logic.py

@@ -1,7 +1,6 @@
 """
 After predictions, we may want to apply some post processing to the predictions
 """
-
 import pandas as pd
 
 
@@ -14,11 +13,10 @@ def clip_predictions_to_minimum_value(
     predictions_df = pd.concat([data, predictions], axis=1)
     # We expect all prediction to be atleast one point improvement
     replace_index = (
-        predictions_df["predictions"]
-        > predictions_df["heat_demand_starting"] - minimum_value
+        predictions_df["sap_starting"] + minimum_value > predictions_df["predictions"]
     )
     predictions_df.loc[replace_index, "predictions"] = (
-        predictions_df.loc[replace_index, "heat_demand_starting"] - minimum_value
+        predictions_df.loc[replace_index, "sap_starting"] + minimum_value
     )
 
     predictions_new = predictions_df["predictions"]

modules/ml-pipeline/src/pipeline/configs/scenarios.yaml

@@ -8,6 +8,6 @@ default:
       # - s3://retrofit-data-dev/scenario_data/27-03-2024-11-38-15/recommendations_scoring_data.parquet
       # - s3://retrofit-data-dev/scenario_data/26-05-2024-08-47-45/recommendations_scoring_data.parquet
       # - s3://retrofit-data-dev/scenario_data/26-05-2024-10-44-53/recommendations_scoring_data.parquet
-      # - s3://retrofit-data-dev/scenario_data/28-05-2024-19-22-41/recommendations_scoring_data.parquet
+      - s3://retrofit-data-dev/scenario_data/28-05-2024-19-22-41/recommendations_scoring_data.parquet
     comparison_output_filepath: ./metrics/scenario_table.md
     metrics_output_filepath: ./metrics/scenario_metrics.md

modules/ml-pipeline/src/pipeline/configs/settings.yaml

@@ -31,13 +31,13 @@ default:
     feature_processor_config:
       subsample_amount: null
       subsample_seed: 0
-      target: heat_demand_ending
+      target: sap_ending
       identifier_columns: ["uprn"]
+      # drop_columns: ["heat_demand_change", "carbon_change", "rdsap_change", "heat_demand_ending", "carbon_ending", "days_to_starting", "days_to_ending"]
       drop_columns: [
-        "heat_demand_change", "carbon_change", "rdsap_change", "sap_ending", "carbon_ending", "days_to_starting", "days_to_ending",
+        "heat_demand_change", "carbon_change", "rdsap_change", "heat_demand_ending", "carbon_ending", "days_to_starting", "days_to_ending",
         'number_habitable_rooms_starting', 'number_habitable_rooms_ending', 'number_heated_rooms_starting', 'number_heated_rooms_ending',
         'number_habitable_rooms', 'number_heated_rooms']
-      # retain_features: ["SAP_STARTING", "TOTAL_FLOOR_AREA_DIFF"]
       retain_features: null
       # retain_features: ['uprn', 'sap_starting', 'hot_water_energy_eff_ending',
       #  'mainheat_energy_eff_ending', 'constituency', 'roof_energy_eff_ending',

modules/ml-pipeline/src/pipeline/dvc.lock

@@ -24,7 +24,7 @@ stages:
         - heat_demand_change
         - carbon_change
         - rdsap_change
-        - sap_ending
+        - heat_demand_ending
         - carbon_ending
         - days_to_starting
         - days_to_ending
@@ -37,7 +37,7 @@ stages:
         default.feature_processor.feature_processor_config.retain_features:
         default.feature_processor.feature_processor_config.subsample_amount:
         default.feature_processor.feature_processor_config.subsample_seed: 0
-        default.feature_processor.feature_processor_config.target: heat_demand_ending
+        default.feature_processor.feature_processor_config.target: sap_ending
         default.feature_processor.feature_processor_type: dataframe
         default.prepare_data.data_filepath:
           s3://retrofit-data-dev/sap_change_model/2024-05-28-19-08-25/dataset_rooms.parquet
@@ -49,8 +49,8 @@ stages:
     outs:
     - path: data/prepared_data/
       hash: md5
-      md5: 13cd955d579de20efe743f82bc434c7e.dir
-      size: 37294025
+      md5: 80c9e138146a1d96b9d16091c207e2e8.dir
+      size: 45056059
       nfiles: 2
   build_model:
     cmd: python 2_build_model.py
@@ -61,8 +61,8 @@ stages:
       size: 4820
     - path: data/prepared_data
       hash: md5
-      md5: 13cd955d579de20efe743f82bc434c7e.dir
-      size: 37294025
+      md5: 80c9e138146a1d96b9d16091c207e2e8.dir
+      size: 45056059
       nfiles: 2
     params:
       configs/build_model.yaml:
@@ -94,18 +94,18 @@ stages:
     outs:
     - path: data/fit_predictions/
       hash: md5
-      md5: b9c9ca64ea6973c409c3a7b8f8ed0c3e.dir
-      size: 2902493
+      md5: d9c9afc05e8780db47c0548b19bf7d19.dir
+      size: 3349989
       nfiles: 1
     - path: data/model/
       hash: md5
-      md5: a9215bba342ed7ec3f97815dfef94e48.dir
-      size: 727501601
+      md5: 13c3100e1486c27a83a8a47491077842.dir
+      size: 773523079
       nfiles: 36
     - path: metrics/fit_metrics.json
       hash: md5
-      md5: 548a431d58cd4f5a3118235dec734372
-      size: 219
+      md5: 2ff70a2a45813e1bcdf2ea3aa8e07d4a
+      size: 224
   generate_predictions:
     cmd: python 3_generate_predictions.py
     deps:
@@ -115,13 +115,13 @@ stages:
       size: 2464
     - path: data/model
       hash: md5
-      md5: a9215bba342ed7ec3f97815dfef94e48.dir
-      size: 727501601
+      md5: 13c3100e1486c27a83a8a47491077842.dir
+      size: 773523079
       nfiles: 36
     - path: data/prepared_data
       hash: md5
-      md5: 13cd955d579de20efe743f82bc434c7e.dir
-      size: 37294025
+      md5: 80c9e138146a1d96b9d16091c207e2e8.dir
+      size: 45056059
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -133,25 +133,25 @@ stages:
     outs:
     - path: data/predictions/
       hash: md5
-      md5: 484781d6b359e458a25e9ab728d6514d.dir
-      size: 380517
+      md5: 5d07bcebf3160a72bb18dfd79106e85c.dir
+      size: 463197
       nfiles: 1
   generate_metrics:
     cmd: python 4_generate_metrics.py
     deps:
     - path: 4_generate_metrics.py
       hash: md5
-      md5: d61bb524f706917f6a3eb72b1ab8bc61
-      size: 3447
+      md5: 4fedb86d89d528f0a6597934ba3890a0
+      size: 3484
     - path: data/predictions
       hash: md5
-      md5: 484781d6b359e458a25e9ab728d6514d.dir
-      size: 380517
+      md5: 5d07bcebf3160a72bb18dfd79106e85c.dir
+      size: 463197
       nfiles: 1
     - path: data/prepared_data
       hash: md5
-      md5: 13cd955d579de20efe743f82bc434c7e.dir
-      size: 37294025
+      md5: 80c9e138146a1d96b9d16091c207e2e8.dir
+      size: 45056059
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -161,8 +161,8 @@ stages:
     outs:
     - path: metrics/metrics.json
       hash: md5
-      md5: 4d246765aff7c45079d02b4d8f7527f7
-      size: 220
+      md5: 3e08df02fd5c5d094bcf936e1338d596
+      size: 223
   generate_scenerio_metrics:
     cmd: python 5_generate_scenarios.py
     deps:
@@ -176,14 +176,15 @@ stages:
           input_dataclient_type: aws-s3
           output_dataclient_type: local
           scenario_data_filepaths:
+          - s3://retrofit-data-dev/scenario_data/28-05-2024-19-22-41/recommendations_scoring_data.parquet
           comparison_output_filepath: ./metrics/scenario_table.md
           metrics_output_filepath: ./metrics/scenario_metrics.md
     outs:
     - path: metrics/scenario_metrics.md
       hash: md5
-      md5: d41d8cd98f00b204e9800998ecf8427e
-      size: 0
+      md5: fa4d6d7bbd7818613800da5f8f37ea96
+      size: 363
     - path: metrics/scenario_table.md
       hash: md5
-      md5: d41d8cd98f00b204e9800998ecf8427e
-      size: 0
+      md5: d6baf100a1623cc2467c2f8221d314c9
+      size: 2133

modules/ml-pipeline/src/pipeline/eda.py

@@ -1,7 +1,6 @@
 """
 Doing some eda on dataset
 """
-
 # Look at response variable
 
 from matplotlib import pyplot as plt
@@ -39,6 +38,7 @@ train_df[[target, "SAP_STARTING"]].plot(y=target, x="SAP_STARTING", style="o")
 train_df[[target, "HEAT_DEMAND_STARTING"]].plot(
     x=target, y="HEAT_DEMAND_STARTING", style="o"
 )
+
 # Both make sense: i.e. the higher the sap, the lower we predict and the higher the heat demand, the higher we predict
 
 # Load the autogluon model and check feature importance
@@ -176,8 +176,6 @@ plot_permutation_importance(exp, fig_kw={"figwidth": 7, "figheight": 6})
 #
 #
 
-from core.MLMetrics import metrics_factory
-
 from core.MLModels import model_factory
 from core.DataClient import dataclient_factory
 import pandas as pd
@@ -218,12 +216,6 @@ mix_df["residual"] = abs(mix_df[predictions_column_name] - mix_df[target])
 mix_df = mix_df.sort_values("residual", ascending=False)
 
 cosine_similarity_df = mix_df[mix_df.columns.difference(["predictions", "residual"])]
-metrics = metrics_factory("Regression")
-metrics.generate_metrics(mix_df["predictions"], mix_df["HEAT_DEMAND_ENDING"])
-
-cosine_similarity_df = mix_df[
-    mix_df.columns.difference(["predictions", "residual", "SAP_ENDING"])
-]
 from sklearn.metrics.pairwise import cosine_similarity
 
 row_index = 0