Update app.py

app.py

@@ -8,67 +8,65 @@ from sklearn.tree import DecisionTreeClassifier
 from sklearn.ensemble import RandomForestClassifier
 from sklearn.metrics import accuracy_score
 
-# 設定 Streamlit 的標題
-st.title('分類模型選擇器')
+# 定義每個分類方法的模型
+def get_classifier(classifier_name):
+    if classifier_name == 'SVM':
+        clf = SVC(kernel='linear', probability=True)
+    elif classifier_name == '堆疊法':
+        estimators = [
+            ('lr', LogisticRegression()),
+            ('rf', RandomForestClassifier()),
+            ('dt', DecisionTreeClassifier())
+        ]
+        clf = StackingClassifier(estimators=estimators, final_estimator=SVC())
+    elif classifier_name == '軟投票':
+        clf1 = LogisticRegression()
+        clf2 = RandomForestClassifier()
+        clf3 = SVC(probability=True)
+        clf = VotingClassifier(estimators=[('lr', clf1), ('rf', clf2), ('svc', clf3)], voting='soft')
+    else:  # 硬投票
+        clf1 = LogisticRegression()
+        clf2 = RandomForestClassifier()
+        clf3 = SVC()
+        clf = VotingClassifier(estimators=[('lr', clf1), ('rf', clf2), ('svc', clf3)], voting='hard')
+    return clf
 
-# 輸入數據集的路徑
-file_path = st.text_input('請輸入數據集的路徑 (CSV 格式)')
+# 加載數據集
+def load_data(file):
+    data = pd.read_csv(file)
+    X = data.iloc[:, :-1].values  # 特徵變量
+    y = data.iloc[:, -1].values   # 目標變量
+    return X, y
 
-# 如果路徑不為空，則加載數據
-if file_path:
-    try:
-        # 加載數據集
-        def load_data(file_path):
-            data = pd.read_csv(file_path)
-            X = data.iloc[:, :-1].values  # 特徵變量
-            y = data.iloc[:, -1].values   # 目標變量
-            return X, y
-        
-        X, y = load_data(file_path)
+# 設定 Streamlit 標題
+st.title("分類模型選擇器")
 
-        # 分割數據集
-        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
+# 上傳 CSV 文件
+uploaded_file = st.file_uploader("上傳 CSV 文件作為數據集", type="csv")
 
-        # 下拉選單讓使用者選擇分類方法
-        classifier_name = st.selectbox(
-            '選擇分類器',
-            ('SVM', '堆疊法', '軟投票', '硬投票')
-        )
+if uploaded_file is not None:
+    # 加載數據集
+    X, y = load_data(uploaded_file)
+    
+    # 分割數據集
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
 
-        # 定義每個分類方法的模型
-        def get_classifier(classifier_name):
-            if classifier_name == 'SVM':
-                clf = SVC(kernel='linear', probability=True)
-            elif classifier_name == '堆疊法':
-                estimators = [
-                    ('lr', LogisticRegression()),
-                    ('rf', RandomForestClassifier()),
-                    ('dt', DecisionTreeClassifier())
-                ]
-                clf = StackingClassifier(estimators=estimators, final_estimator=SVC())
-            elif classifier_name == '軟投票':
-                clf1 = LogisticRegression()
-                clf2 = RandomForestClassifier()
-                clf3 = SVC(probability=True)
-                clf = VotingClassifier(estimators=[('lr', clf1), ('rf', clf2), ('svc', clf3)], voting='soft')
-            else:  # 硬投票
-                clf1 = LogisticRegression()
-                clf2 = RandomForestClassifier()
-                clf3 = SVC()
-                clf = VotingClassifier(estimators=[('lr', clf1), ('rf', clf2), ('svc', clf3)], voting='hard')
-            return clf
+    # 選擇分類器
+    classifier_name = st.selectbox(
+        '選擇分類器',
+        ('SVM', '堆疊法', '軟投票', '硬投票')
+    )
 
-        clf = get_classifier(classifier_name)
+    # 獲取並訓練模型
+    clf = get_classifier(classifier_name)
+    clf.fit(X_train, y_train)
 
-        # 訓練模型
-        clf.fit(X_train, y_train)
+    # 預測結果
+    y_pred = clf.predict(X_test)
 
-        # 預測測試集
-        y_pred = clf.predict(X_test)
+    # 計算準確率
+    acc = accuracy_score(y_test, y_pred)
 
-        # 顯示模型準確率
-        acc = accuracy_score(y_test, y_pred)
-        st.write(f'分類器 = {classifier_name}')
-        st.write(f'準確率 = {acc:.2f}')
-    except Exception as e:
-        st.error(f"加載數據集時出錯: {e}")
+    # 顯示結果
+    st.write(f'分類器 = {classifier_name}')
+    st.write(f'準確率 = {acc:.2f}')