app.py

import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.preprocessing import StandardScaler
from io import BytesIO

# 讓使用者上傳 CSV 檔案
uploaded_file = st.file_uploader("上傳一個 CSV 檔案", type="csv")

if uploaded_file is not None:
    # 讀取上傳的 CSV 檔案
    df = pd.read_csv(uploaded_file)
    
    # 確保數據裡有 "target" 欄位
    if 'target' in df.columns:
        # 準備特徵和目標變量
        X = df.drop('target', axis=1)
        y = df['target']

        # 分割數據
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

        # 標準化特徵
        scaler = StandardScaler()
        X_train_scaled = scaler.fit_transform(X_train)
        X_test_scaled = scaler.transform(X_test)

        # 計算特徵重要性
        def calculate_importance():
            # Linear Regression
            lr = LinearRegression()
            lr.fit(X_train_scaled, y_train)
            lr_importance = np.abs(lr.coef_)

            # CART
            cart = DecisionTreeClassifier(random_state=42)
            cart.fit(X_train, y_train)
            cart_importance = cart.feature_importances_

            # Random Forest
            rf = RandomForestClassifier(n_estimators=100, random_state=42)
            rf.fit(X_train, y_train)
            rf_importance = rf.feature_importances_

            return lr_importance, cart_importance, rf_importance

        # 創建特徵重要性 DataFrame
        lr_importance, cart_importance, rf_importance = calculate_importance()
        feature_importance = pd.DataFrame({
            'Feature': X.columns,
            'Linear Regression': lr_importance,
            'CART': cart_importance,
            'Random Forest': rf_importance
        })

        # 排序
        feature_importance = feature_importance.sort_values('Random Forest', ascending=False)

        # 繪製相關矩陣
        st.write("### 相關矩陣")
        corr_matrix = df.corr()
        plt.figure(figsize=(10, 8))
        sns.heatmap(corr_matrix, annot=True, cmap='coolwarm', fmt='.2f', linewidths=0.5)
        st.pyplot(plt)

        # 分別繪製各個模型的特徵重要性圖表
        def plot_individual_model(model_name):
            plt.figure(figsize=(10, 6))
            plt.bar(feature_importance['Feature'], feature_importance[model_name])
            plt.title(f'{model_name} Feature Importance')
            plt.xlabel('Features')
            plt.ylabel('Importance')
            plt.xticks(rotation=45, ha='right')
            st.pyplot(plt)

        # Streamlit UI
        st.write("### 特徵重要性分析")

        # 分開顯示三個模型的特徵重要性圖表
        st.write("#### Linear Regression")
        plot_individual_model('Linear Regression')

        st.write("#### CART (Decision Tree)")
        plot_individual_model('CART')

        st.write("#### Random Forest")
        plot_individual_model('Random Forest')

        # 顯示數據框
        st.write("### 特徵重要性數據表")
        st.dataframe(feature_importance)

        # 讓使用者下載特徵重要性的 Excel 檔案
        def to_excel(df):
            output = BytesIO()
            writer = pd.ExcelWriter(output, engine='xlsxwriter')
            df.to_excel(writer, index=False, sheet_name='Feature Importance')
            writer.close()  # 使用 close() 來正確保存 Excel 文件
            processed_data = output.getvalue()
            return processed_data

        excel_data = to_excel(feature_importance)
        st.download_button(label='下載特徵重要性數據為 Excel 檔案',
                           data=excel_data,
                           file_name='feature_importance.xlsx',
                           mime='application/vnd.openxmlformats-officedocument.spreadsheetml.sheet')

    else:
        st.error("上傳的檔案中找不到 'target' 欄位，請確認檔案格式。")