Merge pull request #119 from Hestia-Homes/carbon-dev-model

Carbon dev model

.dockerignore

@@ -0,0 +1,9 @@
+modules/ml-pipeline/src/pipeline/data/predictions
+modules/ml-pipeline/src/pipeline/data/fit_predictions
+modules/ml-pipeline/src/pipeline/data/prepared_data
+modules/ml-pipeline/src/pipeline/data/model/allmodels
+modules/ml-pipeline/src/pipeline/metrics
+modules/ml-pipeline/src/pipeline/__pycache__
+modules/ml-pipeline/src/pipeline/.dvc
+modules/ml-pipeline/src/pipeline/analysis
+modules/ml-pipeline/src/pipeline/metrics

.github/workflows/Deploy.yml

@@ -19,8 +19,8 @@ jobs:
 
       - name: Install Serverless and plugins
         run: |
-          npm install -g serverless
+          npm install -g serverless@^3.38.0
-          npm install -g serverless-domain-manager
+          npm install -g serverless-domain-manager@^7.3.8
 
       - name: Install DVC
         run: |

.github/workflows/MLPipelinePullRequest.yml

@@ -98,6 +98,16 @@ jobs:
         git fetch --depth=1 origin ${TARGET_BRANCH}:${TARGET_BRANCH}
         dvc metrics diff --md --all ${TARGET_BRANCH} >> report.md
 
+        echo "## Scenario comparison" >> report.md
+
+        cat metrics/scenario_table.md >> report.md
+
+        echo "" >> report.md
+
+        echo "## Scenario metrics" >> report.md
+
+        cat metrics/scenario_metrics.md >> report.md
+
         cml comment create report.md
 
         # echo "## Residuals plot from model" >> report.md

MODEL_REGISTRY.md

@@ -8,17 +8,17 @@
         "active": true
     },
     "sap": {
-        "version": "v0.11.0",
+        "version": "v0.14.0",
         "stage": {
-            "dev": "v0.11.0"
+            "dev": "v0.14.0"
         },
         "registered": true,
         "active": true
     },
     "heat": {
-        "version": "v0.4.0",
+        "version": "v0.5.0",
         "stage": {
-            "dev": "v0.4.0"
+            "dev": "v0.11.0"
         },
         "registered": true,
         "active": true

deployment/.dockerignore

@@ -1,4 +1,9 @@
-modules/ml-pipeline/src/pipeline/data/predictions*
+modules/ml-pipeline/src/pipeline/data/predictions
-modules/ml-pipeline/src/pipeline/data/prepared_data*
+modules/ml-pipeline/src/pipeline/data/fit_predictions
-modules/ml-pipeline/src/pipeline/data/model/allmodels*
+modules/ml-pipeline/src/pipeline/data/prepared_data
-modules/ml-pipeline/src/pipeline/metrics*
+modules/ml-pipeline/src/pipeline/data/model/allmodels
+modules/ml-pipeline/src/pipeline/metrics
+modules/ml-pipeline/src/__pycache__
+modules/ml-pipeline/src/.dvc
+modules/ml-pipeline/src/analysis
+modules/ml-pipeline/src/metrics

modules/ml-pipeline/src/.dockerignore

@@ -1,4 +1,8 @@
-pipeline/data/predictions*
+pipeline/data/predictions
-pipeline/data/prepared_data/train.parquet*
+pipeline/data/fit_predictions
-pipeline/data/model/allmodels*
+pipeline/data/prepared_data/train.parquet
-pipeline/metrics*
+pipeline/data/fit_predictions
+pipeline/data/model/allmodels
+pipeline/metrics
+pipeline/.dvc
+pipeline/analysis

modules/ml-pipeline/src/Prediction.Dockerfile

@@ -1,7 +1,7 @@
 # Dockerfile that can be used to test loading a model to generate a prediction (part of CI/CD flow)
 FROM python:3.10.12-slim
 
-RUN apt-get update && apt-get install -y libgomp1
+RUN apt-get update && apt-get install -y libgomp1 gcc python3-dev
 
 COPY pipeline/requirements/predictions/requirements.txt requirements.txt
 

modules/ml-pipeline/src/pipeline/5_generate_scenarios.py

@@ -0,0 +1,162 @@
+"""
+Fourth part of the pipeline:
+After the model is built and metrics are generated,
+we want to test this model against known scenarios
+"""
+
+import os
+import pandas as pd
+from core.interface.InterfaceModels import MLModel
+from core.interface.InterfaceDataClient import DataClient
+from core.interface.InterfaceMetrics import MLMetrics
+from configs.post_prediction_logic import post_prediction_logic
+from core.DataClient import dataclient_factory
+from core.MLModels import model_factory
+from core.MLMetrics import metrics_factory
+from core.Logger import logger
+from config import settings
+
+logger.info(f"--- Initiate Parameters ---")
+
+RUNTIME_ENVIRONMENT = os.environ.get("RUNTIME_ENVIRONMENT", "local")
+
+client_params = settings.client
+prepare_data_params = settings.prepare_data
+build_model_params = settings.build_model
+generate_predictions_params = settings.generate_predictions
+generate_metrics_params = settings.generate_metrics
+feature_process_params = settings.feature_processor
+scenarios_params = settings.scenarios
+
+model_filepath = build_model_params["model_save_filepath"]
+target = feature_process_params["feature_processor_config"]["target"]
+scenario_data_filepaths = scenarios_params["scenario_data_filepaths"]
+predictions_column_name = generate_predictions_params["predictions_column_name"]
+comparison_output_filepath = scenarios_params["comparison_output_filepath"]
+metrics_output_filepath = scenarios_params["metrics_output_filepath"]
+
+logger.info(f"--- Initiate MLModel ---")
+
+model = model_factory(build_model_params["model_type"])
+
+logger.info(f"--- Initiate DataClient ---")
+
+# Use data client for input and output, as we use dvc to cache later to the cloud
+input_dataclient_type = scenarios_params["input_dataclient_type"]
+input_dataclient = dataclient_factory(
+    dataclient_type=input_dataclient_type,
+    dataclient_config=client_params[input_dataclient_type],
+)
+
+output_dataclient_type = scenarios_params["output_dataclient_type"]
+output_dataclient = dataclient_factory(
+    dataclient_type=output_dataclient_type,
+    dataclient_config=client_params[output_dataclient_type],
+)
+
+logger.info(f"--- Initiate MLMetrics ---")
+
+metrics = metrics_factory(generate_metrics_params["metrics_type"])
+
+
+def generate_scenario_predictions(
+    input_dataclient: DataClient,
+    output_dataclient: DataClient,
+    model: MLModel,
+    metrics: MLMetrics,
+    model_filepath: str,
+    scenario_data_filepaths: list,
+    predictions_column_name: str,
+    comparison_output_filepath: str,
+    metrics_output_filepath: str,
+):
+    """
+    Given the new model, we generate prediction for expected scenarios
+    """
+
+    logger.info("--- Loading Scenario Data ---")
+
+    scenario_data = pd.DataFrame()
+
+    # If we have no scenario data, we can save empty dataframes
+    if scenario_data_filepaths is None:
+        logger.info("No scenario data filepaths provided")
+        output_dataclient.save_data(
+            obj=scenario_data, location=comparison_output_filepath, save_config=None
+        )
+
+        output_dataclient.save_data(
+            obj=scenario_data, location=metrics_output_filepath, save_config=None
+        )
+        return
+
+    # Can have multiple scenario data files
+    for scenario_data_filepath in scenario_data_filepaths:
+        scenario_data = pd.concat(
+            [
+                scenario_data,
+                input_dataclient.load_data(scenario_data_filepath, load_config=None),
+            ]
+        )
+
+    logger.info("--- Loading Model ---")
+
+    model.load_model(model_filepath)
+
+    logger.info("--- Generating Predictions ---")
+
+    predictions = model.predict(
+        data=scenario_data, post_prediction_logic=post_prediction_logic
+    )
+
+    logger.info("--- Generate Scenario Predicted Impact ---")
+
+    predictions_df = pd.DataFrame(predictions)
+    predictions_df.columns = [predictions_column_name]
+
+    scenario_data = pd.concat([scenario_data, predictions_df], axis=1)
+    scenario_data["predicted_impact"] = abs(
+        scenario_data[predictions_column_name] - scenario_data["sap_starting"]
+    )
+
+    logger.info("--- Generate Metrics ---")
+
+    metrics_dict = metrics.generate_metrics(
+        scenario_data["impact"], scenario_data["predicted_impact"]
+    )
+
+    metrics_df = pd.DataFrame(metrics_dict, index=[0]).T.reset_index()
+    metrics_df.columns = ["metric", "value"]
+
+    logger.info("--- Save prediction into metrics ---")
+
+    output_df = scenario_data[["uprn", "id", "impact", "predicted_impact"]]
+
+    output_dataclient.save_data(
+        obj=output_df, location=comparison_output_filepath, save_config=None
+    )
+
+    output_dataclient.save_data(
+        obj=metrics_df, location=metrics_output_filepath, save_config=None
+    )
+
+
+if __name__ == "__main__":
+
+    logger.info(f"--- {__file__} - Start! ---")
+
+    logger.info(f"--- Generate Scenario Predictions ---")
+
+    generate_scenario_predictions(
+        input_dataclient=input_dataclient,
+        output_dataclient=output_dataclient,
+        model=model,
+        metrics=metrics,
+        model_filepath=model_filepath,
+        scenario_data_filepaths=scenario_data_filepaths,
+        predictions_column_name=predictions_column_name,
+        comparison_output_filepath=comparison_output_filepath,
+        metrics_output_filepath=metrics_output_filepath,
+    )
+
+    logger.info(f"--- {__file__} - Complete! ---")

modules/ml-pipeline/src/pipeline/README.md

@@ -37,3 +37,4 @@ Workflow:
     - This experiment will have the corresponding .dvc files for the hashed model and data
 - Use version control as normal
     - git add, git commit etc
+- To revert change, use `git checkout {COMMIT_HASH}`, followed by `git switch -c {NEW_BRANCH_NAME}`

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

@@ -7,6 +7,7 @@ settings = Dynaconf(
         "./configs/settings.yaml",
         "./configs/build_model.yaml",
         "./configs/analysis.yaml",
+        "./configs/scenarios.yaml",
     ],
 )
 

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

@@ -14,9 +14,9 @@ default:
       output_filepath: ./data/model/allmodels/
       problem_type: regression
       eval_metric: mean_squared_error #mean_absolute_error
-      time_limit: 4000
+      time_limit: 1800
       presets: medium_quality
-      excluded_model_types: ['RF', 'FASTAI', 'CAT', 'NN_TORCH', 'KNN', 'XT']
+      excluded_model_types: ['RF', 'CAT', 'NN_TORCH', 'KNN', 'XT']
       infer_limit: 0.05
       infer_limit_batch_size: 10000
       ag_args_ensemble: {'num_folds_parallel': 2}

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

@@ -0,0 +1,13 @@
+default:
+  scenarios:
+    input_dataclient_type: aws-s3
+    output_dataclient_type: local
+    scenario_data_filepaths:
+      # - s3://retrofit-data-dev/scenario_data/22-03-2024-19-20-09/recommendations_scoring_data.parquet
+      # - s3://retrofit-data-dev/scenario_data/24-03-2024-20-23-25/recommendations_scoring_data.parquet
+      # - 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
+    comparison_output_filepath: ./metrics/scenario_table.md
+    metrics_output_filepath: ./metrics/scenario_metrics.md

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

@@ -18,7 +18,10 @@ default:
   prepare_data:
     input_dataclient_type: aws-s3
     output_dataclient_type: local
-    data_filepath: s3://retrofit-data-dev/sap_change_model/2024-03-22-18-56-53/dataset_rooms.parquet
+    # data_filepath: s3://retrofit-data-dev/sap_change_model/2024-03-22-18-56-53/dataset_rooms.parquet
+    # data_filepath: s3://retrofit-data-dev/sap_change_model/2024-05-25-08-36-36/dataset_rooms.parquet
+    # data_filepath: s3://retrofit-data-dev/sap_change_model/2024-05-26-10-31-39/dataset_rooms.parquet
+    data_filepath: s3://retrofit-data-dev/sap_change_model/2024-05-28-19-08-25/dataset_rooms.parquet
     train_proportion: 0.9
     output_train_filepath: ./data/prepared_data/train.parquet
     output_test_filepath: ./data/prepared_data/test.parquet
@@ -37,6 +40,29 @@ default:
         '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',
+      #  'walls_energy_eff_ending', 'secondheat_description_ending',
+      #  'property_type', 'mainheatc_energy_eff_ending', 'built_form',
+      #  'walls_insulation_thickness_ending', 'potential_energy_efficiency',
+      #  'transaction_type_ending',
+      #  'floor_thermal_transmittance_ending',
+      #  'low_energy_lighting_ending', 'heat_demand_starting',
+      #  'photo_supply_ending', 'carbon_starting',
+      #  'walls_thermal_transmittance_ending',
+      #  'roof_insulation_thickness_ending',
+      #  'total_floor_area_ending', 'number_open_fireplaces_ending',
+      #  'windows_energy_eff_ending',
+      #  'floor_height_ending',
+      #  'extension_count_ending',
+      #  'has_air_source_heat_pump_ending',
+      #  'charging_system_ending', 'construction_age_band', 'glazed_type_ending',
+      #  'roof_thermal_transmittance_ending',
+      #  'floor_insulation_thickness_ending', 'has_mains_gas_ending',
+      #  'estimated_perimeter_starting', 'energy_consumption_potential',
+      #  'environment_impact_potential', 'heater_type_ending',
+      #  'multi_glaze_proportion_ending',
+      #  'lighting_energy_eff_ending', 'fixed_lighting_outlets_count']
 
   generate_predictions:
     input_dataclient_type: local

modules/ml-pipeline/src/pipeline/core/DataClient.py

@@ -245,7 +245,8 @@ class LocalClient:
 
         save_methods = {
             ".parquet": self._save_parquet,
-            ".json": self._save_json
+            ".json": self._save_json,
+            ".md": self._save_md,
             # "": _save_directory(**save_config),
             # ADD MORE save_methods HERE
         }
@@ -294,3 +295,10 @@ class LocalClient:
         # Write the contents of the buffer to the local file
         with open(location, "wb") as f:
             f.write(buffer.getvalue())
+
+    def _save_md(self, obj: pd.DataFrame, location: str, save_config: dict):
+        """
+        Save object as markdown
+        """
+
+        obj.to_markdown(location, **save_config)

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

@@ -39,8 +39,8 @@ stages:
         default.feature_processor.feature_processor_config.subsample_seed: 0
         default.feature_processor.feature_processor_config.target: carbon_ending
         default.feature_processor.feature_processor_type: dataframe
-        default.prepare_data.data_filepath:
+        default.prepare_data.data_filepath: 
-          s3://retrofit-data-dev/sap_change_model/2024-03-22-18-56-53/dataset_rooms.parquet
+          s3://retrofit-data-dev/sap_change_model/2024-05-28-19-08-25/dataset_rooms.parquet
         default.prepare_data.input_dataclient_type: aws-s3
         default.prepare_data.output_dataclient_type: local
         default.prepare_data.output_test_filepath: ./data/prepared_data/test.parquet
@@ -49,8 +49,8 @@ stages:
     outs:
     - path: data/prepared_data/
       hash: md5
-      md5: 824541f44e6538d2ef10e9d754c79743.dir
+      md5: e2efac20634b919381adfb962a42d40a.dir
-      size: 36691842
+      size: 36961727
       nfiles: 2
   build_model:
     cmd: python 2_build_model.py
@@ -61,8 +61,8 @@ stages:
       size: 4820
     - path: data/prepared_data
       hash: md5
-      md5: 824541f44e6538d2ef10e9d754c79743.dir
+      md5: e2efac20634b919381adfb962a42d40a.dir
-      size: 36691842
+      size: 36961727
       nfiles: 2
     params:
       configs/build_model.yaml:
@@ -79,11 +79,10 @@ stages:
               output_filepath: ./data/model/allmodels/
               problem_type: regression
               eval_metric: mean_squared_error
-              time_limit: 4000
+              time_limit: 1800
               presets: medium_quality
               excluded_model_types:
               - RF
-              - FASTAI
               - CAT
               - NN_TORCH
               - KNN
@@ -95,18 +94,18 @@ stages:
     outs:
     - path: data/fit_predictions/
       hash: md5
-      md5: 5a3091120d3497fa00b994d91bc7e5eb.dir
+      md5: d2568a3244df4d3444b6190599f74b96.dir
-      size: 3664806
+      size: 3661106
       nfiles: 1
     - path: data/model/
       hash: md5
-      md5: 074da8dcfa515b9f3d082b21c7d76616.dir
+      md5: 756100e033e0bd4445a437e43f4c53af.dir
-      size: 721558897
+      size: 730442848
-      nfiles: 31
+      nfiles: 36
     - path: metrics/fit_metrics.json
       hash: md5
-      md5: 728a49dcef5a98182325df455f929a33
+      md5: 3bcb3b9728521cd341eb71af109ca778
-      size: 225
+      size: 227
   generate_predictions:
     cmd: python 3_generate_predictions.py
     deps:
@@ -116,13 +115,13 @@ stages:
       size: 2464
     - path: data/model
       hash: md5
-      md5: 074da8dcfa515b9f3d082b21c7d76616.dir
+      md5: 756100e033e0bd4445a437e43f4c53af.dir
-      size: 721558897
+      size: 730442848
-      nfiles: 31
+      nfiles: 36
     - path: data/prepared_data
       hash: md5
-      md5: 824541f44e6538d2ef10e9d754c79743.dir
+      md5: e2efac20634b919381adfb962a42d40a.dir
-      size: 36691842
+      size: 36961727
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -134,8 +133,8 @@ stages:
     outs:
     - path: data/predictions/
       hash: md5
-      md5: 680f51234d214d4cab9e6a064c75fc5d.dir
+      md5: 09f3584d6fbd447dd2714eb2774139d5.dir
-      size: 499546
+      size: 499683
       nfiles: 1
   generate_metrics:
     cmd: python 4_generate_metrics.py
@@ -146,13 +145,13 @@ stages:
       size: 3484
     - path: data/predictions
       hash: md5
-      md5: 680f51234d214d4cab9e6a064c75fc5d.dir
+      md5: 09f3584d6fbd447dd2714eb2774139d5.dir
-      size: 499546
+      size: 499683
       nfiles: 1
     - path: data/prepared_data
       hash: md5
-      md5: 824541f44e6538d2ef10e9d754c79743.dir
+      md5: e2efac20634b919381adfb962a42d40a.dir
-      size: 36691842
+      size: 36961727
       nfiles: 2
     params:
       configs/settings.yaml:
@@ -162,5 +161,29 @@ stages:
     outs:
     - path: metrics/metrics.json
       hash: md5
-      md5: 67b7ab30a4b0839d20bc6eb0c84e4dd1
+      md5: abf8720d06f073f47501aa1172527e9e
-      size: 226
+      size: 225
+  generate_scenerio_metrics:
+    cmd: python 5_generate_scenarios.py
+    deps:
+    - path: 5_generate_scenarios.py
+      hash: md5
+      md5: 40506749fefd926d47c60ff5b16db307
+      size: 5337
+    params:
+      configs/scenarios.yaml:
+        default.scenarios:
+          input_dataclient_type: aws-s3
+          output_dataclient_type: local
+          scenario_data_filepaths:
+          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
+    - path: metrics/scenario_table.md
+      hash: md5
+      md5: d41d8cd98f00b204e9800998ecf8427e
+      size: 0

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

@@ -71,6 +71,17 @@ stages:
     outs:
     - metrics/metrics.json
     always_changed: true
+  generate_scenerio_metrics:
+    cmd: python 5_generate_scenarios.py
+    deps:
+    - 5_generate_scenarios.py
+    params:
+    - configs/scenarios.yaml:
+      - default.scenarios
+    outs:
+    - metrics/scenario_table.md
+    - metrics/scenario_metrics.md
+    always_changed: true
 metrics:
   - metrics/metrics.json
   - metrics/fit_metrics.json

modules/ml-pipeline/src/pipeline/metrics/.gitignore

@@ -1,2 +1,4 @@
 /fit_metrics.json
 /metrics.json
+/scenario_table.md
+/scenario_metrics.md

modules/ml-pipeline/src/pipeline/requirements/predictions/requirements-dev.txt

@@ -1,7 +1,7 @@
 joblib==1.3.2
 boto3==1.28.17
 pandas==2.1.4
-autogluon==1.0.0
+autogluon.tabular[all]==1.0.0
 dynaconf==3.2.1
 pyarrow==13.0.0
 pre-commit==3.3.3

modules/ml-pipeline/src/pipeline/requirements/predictions/requirements.txt

@@ -1,7 +1,7 @@
 joblib==1.3.2
 boto3==1.28.17
 pandas==2.1.4
-autogluon==1.0.0
+autogluon.tabular[all]==1.0.0
 dynaconf==3.2.1
 pyarrow==13.0.0
 PyYAML==6.0.1

modules/ml-pipeline/src/pipeline/requirements/training/requirements-dev.txt

@@ -1,7 +1,7 @@
 joblib==1.3.2
 boto3==1.28.17
 pandas==2.1.4
-autogluon==1.0.0
+autogluon.tabular[all]==1.0.0
 ray==2.6.3
 dynaconf==3.2.1
 alibi==0.9.5

modules/ml-pipeline/src/pipeline/requirements/training/requirements.txt

@@ -1,4 +1,4 @@
 boto3==1.28.41
 pandas==2.1.4
-autogluon==1.0.0
+autogluon.tabular[all]==1.0.0
 dynaconf==3.2.1

modules/ml-pipeline/src/pipeline/requirements/version_control/requirements.txt

@@ -1,4 +1,4 @@
-dvc==3.36.0
+dvc==3.51.0
-dvc-s3==3.0.1
+dvc-s3==3.2.0
-gto==1.6.1
+gto==1.7.1
 pyOpenSSL==23.3.0