Phase 1 κ²°κ³Όλ₯Ό μκ°ννλ λ‘컬 Streamlit λμ보λ ꡬμΆ
EcoTracing/
βββ streamlit/
βββ app.py # λ©μΈ μ§μ
μ , ννλ©΄
βββ utils/
β βββ __init__.py # ν¨ν€μ§ μΈμ
β βββ loader.py # config/λͺ¨λΈ/κ²°κ³Ό νμΌ λ‘λ
β βββ predictor.py # μλμ§ κ³΅μ κ³μ° + λͺ¨λΈ μμΈ‘
βββ pages/
βββ predictor.py # CPU/Mem μ¬λΌμ΄λ -> μ€μκ° μλμ§ μμΈ‘
βββ eda.py # EDA μκ°ν (λΆν¬, μκ°λ ν¨ν΄, μ°μ λ)
βββ model_compare.py # 4κ° λͺ¨λΈ μ±λ₯ λΉκ΅ μ°¨νΈ
βββ feature_importance.py # νΌμ² μ€μλ λ°/νμ΄ μ°¨νΈ
| νμ΄μ§ | λ΄μ© |
|---|---|
| π ν (app.py) | νλ‘μ νΈ μκ°, Phase 1 ν΅μ¬ κ²°κ³Ό μμ½ |
| π Predictor | CPU/Memory μ¬λΌμ΄λ μ λ ₯ β μ€μκ° μλμ§ μμΈ‘ + νμ λ°°μΆλ + μ§κ΄μ λΉκ΅ |
| π EDA | CPU/Memory λΆν¬, μκ°λλ³ ν¨ν΄, μ°μ λ |
| π Model Compare | 4κ° λͺ¨λΈ RMSE/R2/MAPE/νμ΅μκ° λΉκ΅ μ°¨νΈ |
| π Feature Importance | νΌμ² μ€μλ λ°/νμ΄ μ°¨νΈ + ν΄μ |
paths:
processed_file: "instance_usage_full_processed.parquet"
model:
model_names:
lightgbm: "energy_model_lightgbm.pkl"
xgboost: "energy_model_xgboost.pkl"
randomforest: "energy_model_randomforest.pkl"
catboost: "energy_model_catboost.pkl"
results:
results_json: "phase1_full_results.json"
feature_importance_csv: "feature_importance_comparison.csv"
cd C:\\Git\\EcoTracing
streamlit run streamlit/app.py