Anomaly number of UV oven lamps prediction.
-
Setup
mkdir processed,mkdir output -
Copy train1 data from readonly directory
/TOPIC/projectA/to./data/raw/train1/
python -m check_tr1 --data-dir /TOPIC/projectA/(run for competition environment only) -
Generate fake data (train2 and test) using train1 data
python -m gen_fake_data --data-dir ./data/raw/train1/ -
Process train1 data
python -m dp --data-type tr1 -
Train models using processed train1 data with 10 seeds
python -m train_eval --proc-file train_1.csv --n-seeds 10
-
Check train2 data format
python -m check_tr2 --data-dir ./data/fake_projectA/ -
Process train1&train2 data
python -m dp --data-type tr2 -
Train models using processed train1&train2 data with 10 seeds
python -m train_eval --proc-file train_1_2.csv --n-seeds 10
- Process test data
python -m dp --data-type ts --ts-source tr_12 - Inference
python -m infer --model-dir ./output/fold --save-dir ./output/ --save-fname 111011_projectA_ans.csv
├── check_tr1.py
├── check_tr2.py
├── config
│ └── model
│ └── lgbm_template.yaml
├── data
│ ├── fake_projectA
│ │ ├── test
│ │ │ └── accumulation_hour3.csv
│ │ └── train2
│ │ ├── accumulation_hour2.csv
│ │ └── anomaly_train2.csv
│ ├── processed
│ │ ├── feat_cols.csv
│ │ ├── oid2idx.pkl
│ │ ├── test.csv
│ │ ├── train_1_2.csv
│ │ └── train_1.csv
│ └── raw
│ ├── test
│ │ ├── accumulation_hour3.csv
│ │ ├── cooler.csv
│ │ └── projectA_template.csv
│ ├── train1
│ │ ├── accumulation_hour1.csv
│ │ ├── anomaly_train1.csv
│ │ ├── A_raw_data.zip
│ │ ├── cooler.csv
│ │ └── power.csv
│ └── train2
│ ├── accumulation_hour2.csv
│ └── anomaly_train2.csv
├── dp.py
├── gen_fake_data.py
├── infer.py
├── modeling
│ └── build.py
├── notebooks
│ └── simple_eda.ipynb
├── output
│ ├── 111011_projectA_ans.csv
│ └── fold
| ├── seed0_fold0.pkl
│ ├── seed0_fold1.pkl
│ ├── seed0_fold2.pkl
│ ├── seed0_fold3.pkl
│ ├── seed0_fold4.pkl
│ ├── seed0_fold5.pkl
│ ├── seed1_fold0.pkl
│ ├── seed1_fold1.pkl
│ ├── seed1_fold2.pkl
│ ├── seed1_fold3.pkl
│ ├── seed1_fold4.pkl
| ...
├── pyproject.toml
├── README.md
├── setup.cfg
├── tools
│ └── train_eval.py
├── train_eval.py
└── utils
└── utils.py