junte/ML
1
0
Fork 0
mirror of https://github.com/Hestia-Homes/ML.git synced 2026-06-08 11:17:25 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

optimise to squared error to penalise large errors

Browse source
This commit is contained in:
Michael Duong 2023-09-29 15:53:12 +00:00
parent f67b138406
commit c0d73d8b9e
4 changed files with 31 additions and 26 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
modules/ml-pipeline/src/pipeline/configs/build_model.yaml
View file

@ -12,7 +12,7 @@ default:
AutogluonAutoML:
output_filepath: ./data/model/autogluonmodel/
problem_type: regression
eval_metric: mean_absolute_error
eval_metric: mean_squared_error #mean_absolute_error
time_limit: 1000
presets: medium_quality
excluded_model_types: ['KNN']
excluded_model_types: ['KNN', 'RF']

2
modules/ml-pipeline/src/pipeline/configs/feature_processor_logic.py
View file

@ -27,7 +27,7 @@ def remove_starting_columns(df):
# return df
business_logic = {
"remove_starting_columns": remove_starting_columns
# "remove_starting_columns": remove_starting_columns
# "keep_ENDING_COLUMNS": keep_ending_columns
}

6
modules/ml-pipeline/src/pipeline/configs/settings.yaml
View file

@ -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/dataset_without_differencing.parquet
# data_filepath: s3://retrofit-data-dev/sap_change_model/dataset_with_differencing.parquet
# data_filepath: s3://retrofit-data-dev/sap_change_model/floor_area_clean_test.parquet
# data_filepath: s3://retrofit-data-dev/sap_change_model/dataset_without_differencing.parquet
data_filepath: s3://retrofit-data-dev/sap_change_model/dataset.parquet
train_proportion: 0.9
output_train_filepath: ./data/prepared_data/train.parquet
output_test_filepath: ./data/prepared_data/test.parquet
@ -30,6 +33,7 @@ default:
subsample_seed: 0
target: SAP_ENDING
drop_columns: ["UPRN", "HEAT_DEMAND_CHANGE", "CARBON_CHANGE", "RDSAP_CHANGE", "HEAT_DEMAND_ENDING", "CARBON_ENDING"]
# retain_features: ["SAP_STARTING", "TOTAL_FLOOR_AREA_DIFF"]
retain_features: null
generate_predictions:

45
modules/ml-pipeline/src/pipeline/dvc.lock
View file

@ -21,7 +21,7 @@ stages:
default.feature_processor.feature_processor_config.subsample_seed: 0
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/dataset_without_differencing.parquet
default.prepare_data.data_filepath: s3://retrofit-data-dev/sap_change_model/dataset.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
@ -30,8 +30,8 @@ stages:
outs:
- path: data/prepared_data/
hash: md5
md5: f9ef7ad073b43b249b43faa75c62fe07.dir
size: 21115444
md5: 951ad046d4fca2b977a314f9520e8235.dir
size: 28249626
nfiles: 2
build_model:
cmd: python 2_build_model.py
@ -42,8 +42,8 @@ stages:
size: 5181
- path: data/prepared_data
hash: md5
md5: f9ef7ad073b43b249b43faa75c62fe07.dir
size: 21115444
md5: 951ad046d4fca2b977a314f9520e8235.dir
size: 28249626
nfiles: 2
params:
configs/build_model.yaml:
@ -58,21 +58,22 @@ stages:
AutogluonAutoML:
output_filepath: ./data/model/autogluonmodel/
problem_type: regression
eval_metric: mean_absolute_error
eval_metric: mean_squared_error
time_limit: 1000
presets: medium_quality
excluded_model_types:
- KNN
- RF
outs:
- path: data/model/
hash: md5
md5: d073af40ba5c7c2d9b8064665062f51e.dir
size: 363710367
md5: 6cedf54b03ce278149c67ed118ed0c59.dir
size: 408028954
nfiles: 20
- path: metrics/fit_metrics.json
hash: md5
md5: dcd9ea03a2771077e1bd14018bb7fd18
size: 183
md5: 6352c43a896bedc387ac6fb0e1c6e09f
size: 184
generate_predictions:
cmd: python 3_generate_predictions.py
deps:
@ -82,13 +83,13 @@ stages:
size: 4720
- path: data/model
hash: md5
md5: d073af40ba5c7c2d9b8064665062f51e.dir
size: 363710367
md5: 6cedf54b03ce278149c67ed118ed0c59.dir
size: 408028954
nfiles: 20
- path: data/prepared_data
hash: md5
md5: f9ef7ad073b43b249b43faa75c62fe07.dir
size: 21115444
md5: 951ad046d4fca2b977a314f9520e8235.dir
size: 28249626
nfiles: 2
params:
configs/settings.yaml:
@ -100,8 +101,8 @@ stages:
outs:
- path: data/predictions/
hash: md5
md5: a2ecfae1e418fe9cb9fe044c148bbb37.dir
size: 381538
md5: 5f5b251bb260c366b73b9481b8ffec4c.dir
size: 374624
nfiles: 1
generate_metrics:
cmd: python 4_generate_metrics.py
@ -112,13 +113,13 @@ stages:
size: 4487
- path: data/predictions
hash: md5
md5: a2ecfae1e418fe9cb9fe044c148bbb37.dir
size: 381538
md5: 5f5b251bb260c366b73b9481b8ffec4c.dir
size: 374624
nfiles: 1
- path: data/prepared_data
hash: md5
md5: f9ef7ad073b43b249b43faa75c62fe07.dir
size: 21115444
md5: 951ad046d4fca2b977a314f9520e8235.dir
size: 28249626
nfiles: 2
params:
configs/settings.yaml:
@ -128,8 +129,8 @@ stages:
outs:
- path: metrics/metrics.json
hash: md5
md5: ec02774fd01243fa4706189c60087ccf
size: 182
md5: 1268ce542c45da1f38286aa941f21c1e
size: 183
startup_cleanup:
cmd: python 0_startup_cleanup.py
deps: