| import streamlit as st | |
| from pycaret.regression import * | |
| from pathlib import Path | |
| path_root = Path(Path.cwd()) | |
| from pycaret.datasets import get_data | |
| st.markdown("# Regression❄️") | |
| with st.sidebar: | |
| st.sidebar.markdown("# Regression Feature ❄️") | |
| evaluation_option = st.selectbox("evaluation", ["residuals", "error", "cooks", "rfe", "learning", "vc", "manifold", "feature", "feature_all", "residuals_interactive", "parameter", "tree"]) | |
| data = get_data('insurance') | |
| s = setup(data, target = 'charges') | |
| st.markdown("## Compare model") | |
| best = compare_models() | |
| df_metric = pull() | |
| st.dataframe(df_metric) | |
| st.markdown("## Evaluation") | |
| evaluate_model(best) | |
| image_options = {"error": "Prediction Error.png", | |
| "residuals": "Residuals.png", | |
| "cooks": "unknow", | |
| "rfe": "unknow", | |
| "learning": "Learning Curve.png", | |
| "vc": "Validation Curve.png", | |
| "manifold": "Manifold Curve.png", | |
| "feature": "Feature Importance.png", | |
| "feature_all": "Feature Importance (All).png", | |
| "residuals_interactive": "unknow" | |
| } | |
| try: | |
| st.markdown(f"## {evaluation_option}") | |
| plot_model(best, plot = evaluation_option, save = 'images') | |
| st.image(str(path_root.joinpath(f"images/{image_options[evaluation_option]}"))) | |
| except e: | |
| st.text(e) | |