Update app.py

app.py

@@ -1,18 +1,27 @@
-# 如果在 Colab，本區塊可先執行一次（本機已安裝就略過）
-!pip -q install scikit-learn pandas matplotlib gradio
+# -*- coding: utf-8 -*-
+"""
+Hugging Face Spaces / Gradio app
+步驟導引：資料 → 切分/訓練 → 評估(Accuracy+混淆矩陣) → 測試介面
+資料欄位需為: text, label (label 為 0/1)
+"""
 
+import io
+import numpy as np
 import pandas as pd
+import matplotlib.pyplot as plt
+
+import gradio as gr
 from sklearn.model_selection import train_test_split
+from sklearn.pipeline import Pipeline
 from sklearn.feature_extraction.text import CountVectorizer, TfidfTransformer
 from sklearn.linear_model import LogisticRegression
-from sklearn.naive_bayes import MultinomialNB
-from sklearn.pipeline import Pipeline
 from sklearn.metrics import accuracy_score, classification_report, confusion_matrix
-import matplotlib.pyplot as plt
-import numpy as np
 
-# 1) 準備一個小型「示範資料集」（你也可以改成班級自建）
-data = [
+
+# -------------------------
+# 內建示範資料（30 筆）
+# -------------------------
+SAMPLE_DATA = [
     ("這家餐廳超好吃，我下次還要來！", 1),
     ("服務人員很親切，體驗很棒", 1),
     ("環境乾淨、氣氛舒服，推！", 1),
@@ -44,61 +53,216 @@ data = [
     ("週年限定套餐超值，朋友都喜歡", 1),
     ("衛生紙與餐具不足，需要一直跟店員拿", 0),
 ]
+SAMPLE_DF = pd.DataFrame(SAMPLE_DATA, columns=["text", "label"])
+
 
+# -------------------------
+# 工具函式
+# -------------------------
+def read_csv_file(file_obj) -> pd.DataFrame:
+    """從上傳檔案讀 CSV"""
+    return pd.read_csv(file_obj, encoding_errors="ignore")
 
-df = pd.DataFrame(data, columns=["text", "label"])
 
+def read_csv_text(text_block: str) -> pd.DataFrame:
+    """從貼上的 CSV 文字讀取（需含表頭 text,label）"""
+    return pd.read_csv(io.StringIO(text_block))
 
-# 2) 訓練/測試 切分
-X_train, X_test, y_train, y_test = train_test_split(
-    df["text"], df["label"], test_size=0.3, random_state=42, stratify=df["label"]
-)
 
-# 3) 做一條「處理流水線」：文字 -> 向量化 -> TF-IDF -> 模型
-#    (A) 邏輯斯迴歸版本（入門也很好用，常見基線）
-logreg_clf = Pipeline([
-    # 中文建議用「字 n-gram」，避免斷詞問題；2~4字能抓到詞片段
-    ("vect", CountVectorizer(analyzer="char", ngram_range=(2, 4))),
-    ("tfidf", TfidfTransformer()),
-    ("clf", LogisticRegression(max_iter=200))
-])
+def check_df(df: pd.DataFrame) -> str:
+    """基本欄位檢查，回傳錯誤訊息（無錯回空字串）"""
+    cols = set(df.columns.str.lower())
+    need = {"text", "label"}
+    if not need.issubset(cols):
+        return "❌ CSV 需要兩個欄位：text, label（0/1）。"
+    if df.empty:
+        return "❌ 資料為空。"
+    return ""
 
-logreg_clf.fit(X_train, y_train)
-y_pred_lr = logreg_clf.predict(X_test)
 
-print("=== Logistic Regression ===")
-print("Accuracy:", accuracy_score(y_test, y_pred_lr))
-print(classification_report(y_test, y_pred_lr, digits=3))
+def build_pipeline() -> Pipeline:
+    """字 n-gram + TF-IDF + Logistic Regression"""
+    return Pipeline([
+        ("vect", CountVectorizer(analyzer="char", ngram_range=(2, 4))),
+        ("tfidf", TfidfTransformer()),
+        ("clf", LogisticRegression(max_iter=200)),
+    ])
 
-# 5) 畫一張混淆矩陣（看錯在哪）
-def plot_cm(y_true, y_pred, title):
+
+def plot_confusion_matrix(y_true, y_pred, labels=("負面(0)", "正面(1)")):
+    """回傳 matplotlib 圖物件（Gradio 會顯示）"""
     cm = confusion_matrix(y_true, y_pred)
-    fig, ax = plt.subplots(figsize=(4,3))
-    im = ax.imshow(cm)
-    ax.set_title(title)
+    fig, ax = plt.subplots(figsize=(4, 3))
+    im = ax.imshow(cm, cmap="Blues")
+    ax.set_title("Confusion Matrix - Logistic Regression")
     ax.set_xlabel("Predicted")
     ax.set_ylabel("True")
-    ax.set_xticks([0,1]); ax.set_yticks([0,1])
-    ax.set_xticklabels(["負面(0)","正面(1)"]); ax.set_yticklabels(["負面(0)","正面(1)"])
-    # 印數字
+    ax.set_xticks([0, 1]); ax.set_yticks([0, 1])
+    ax.set_xticklabels(labels); ax.set_yticklabels(labels)
     for i in range(cm.shape[0]):
         for j in range(cm.shape[1]):
-            ax.text(j, i, cm[i, j], ha="center", va="center")
-    plt.show()
+            ax.text(j, i, cm[i, j], ha="center", va="center", color="black")
+    fig.colorbar(im, ax=ax, fraction=0.046, pad=0.04)
+    fig.tight_layout()
+    return fig
 
-plot_cm(y_test, y_pred_lr, "Confusion Matrix - Logistic Regression")
 
+# -------------------------
+# Gradio 回呼：Step1 載入資料
+# -------------------------
+def use_sample_dataset():
+    df = SAMPLE_DF.copy()
+    msg = f"✅ 已載入內建示範資料，共 {len(df)} 筆。"
+    return df, df.head(10), msg
 
-# 6) 建立界面
-import gradio as gr
 
-with gr.Blocks(title="情感分類小幫手（LogReg + 字 n-gram）") as demo:
-    gr.Markdown("### 輸入一句中文，AI 幫你判斷正/負面。")
-    inp = gr.Textbox(label="輸入文字", lines=2, placeholder="例：這家好吃到爆！")
-    submit = gr.Button("Submit")
-    out = gr.Textbox(label="AI 判斷結果", interactive=False)
+def load_from_file(file):
+    try:
+        df = read_csv_file(file)
+        err = check_df(df)
+        if err:
+            return gr.State(None), None, err
+        df = df.rename(columns={"Text": "text", "Label": "label"})
+        msg = f"✅ 已載入檔案：{len(df)} 筆。"
+        return df, df.head(10), msg
+    except Exception as e:
+        return gr.State(None), None, f"❌ 載入失敗：{e}"
+
+
+def load_from_text(csv_text):
+    try:
+        df = read_csv_text(csv_text)
+        err = check_df(df)
+        if err:
+            return gr.State(None), None, err
+        df = df.rename(columns={"Text": "text", "Label": "label"})
+        msg = f"✅ 已載入貼上資料：{len(df)} 筆。"
+        return df, df.head(10), msg
+    except Exception as e:
+        return gr.State(None), None, f"❌ 解析失敗：{e}"
+
+
+# -------------------------
+# Gradio 回呼：Step2 訓練與評估
+# -------------------------
+def train_and_eval(df, test_ratio):
+    if df is None:
+        return None, "❌ 請先載入資料。", None, None
+
+    # 乾淨處理
+    df = df.dropna(subset=["text", "label"]).copy()
+    df["label"] = df["label"].astype(int)
+
+    # 切分
+    X_train, X_test, y_train, y_test = train_test_split(
+        df["text"], df["label"],
+        test_size=test_ratio, random_state=42, stratify=df["label"]
+    )
+
+    # 訓練
+    model = build_pipeline()
+    model.fit(X_train, y_train)
+
+    # 評估
+    y_pred = model.predict(X_test)
+    acc = accuracy_score(y_test, y_pred)
+    rpt = classification_report(y_test, y_pred, digits=3)
+    rpt_text = f"Accuracy: {acc:.3f}\n\n{rpt}"
+
+    # 圖
+    fig = plot_confusion_matrix(y_test, y_pred)
+
+    # 類別分布小提示
+    dist = df["label"].value_counts().sort_index().to_dict()
+    tip = f"資料分布：0→{dist.get(0,0)} , 1→{dist.get(1,0)}"
+
+    return model, rpt_text + "\n" + tip, fig, "✅ 訓練完成，可以到下一步做測試。"
+
+
+# -------------------------
+# Gradio 回呼：Step3 推論
+# -------------------------
+def predict_one(text, model):
+    if model is None:
+        return "❌ 尚未訓練模型。"
+    if not text or not text.strip():
+        return "請先輸入文字。"
+    proba = model.predict_proba([text])[0][1]
+    label = "正面(1)" if proba >= 0.5 else "負面(0)"
+    return f"{label}（信心 {proba:.2f}）"
+
+
+def predict_batch(df_texts, model):
+    if model is None:
+        return None
+    if df_texts is None or df_texts.empty:
+        return None
+    texts = df_texts.iloc[:, 0].astype(str).tolist()
+    probas = model.predict_proba(texts)[:, 1]
+    labels = (probas >= 0.5).astype(int)
+    out = pd.DataFrame({"text": texts, "prob_pos": probas, "pred": labels})
+    return out
+
+
+# -------------------------
+# 介面
+# -------------------------
+with gr.Blocks(title="情感分類小幫手（LogReg + 字 n-gram）", theme=gr.themes.Soft()) as demo:
+    gr.Markdown("## 情感分類小幫手（LogReg + 字 n-gram）\n一步一步完成：資料 → 訓練 → 評估 → 測試")
+
+    state_df = gr.State()      # 保存目前資料集
+    state_model = gr.State()   # 保存已訓練模型
+
+    with gr.Tab("Step 1｜載入資料"):
+        gr.Markdown(
+            "上傳 **CSV (text,label)** 或貼上文字，或使用內建示範資料。\n"
+            "- label：0=負面、1=正面\n"
+            "- CSV 必須含表頭：`text,label`\n"
+        )
+        with gr.Row():
+            file_in = gr.File(label="上傳 CSV")
+            btn_load_file = gr.Button("讀取檔案")
+        with gr.Row():
+            txt_in = gr.Textbox(lines=5, label="或貼上 CSV 文字（需含表頭 text,label）")
+            btn_load_text = gr.Button("讀取貼上文字")
+        btn_use_sample = gr.Button("使用內建示範資料（30筆）")
+
+        msg1 = gr.Markdown()
+        df_preview = gr.Dataframe(headers=["text", "label"], label="資料預覽（前10筆）", interactive=False)
+
+        btn_use_sample.click(fn=use_sample_dataset, outputs=[state_df, df_preview, msg1])
+        btn_load_file.click(fn=load_from_file, inputs=[file_in], outputs=[state_df, df_preview, msg1])
+        btn_load_text.click(fn=load_from_text, inputs=[txt_in], outputs=[state_df, df_preview, msg1])
+
+    with gr.Tab("Step 2｜切分與訓練"):
+        gr.Markdown("設定測試集比例，按下 **開始訓練**。")
+        split = gr.Slider(0.1, 0.5, value=0.3, step=0.05, label="測試集比例 test_size")
+        btn_train = gr.Button("開始訓練")
+        train_msg = gr.Markdown()
+        report_box = gr.Textbox(label="評估報告（Accuracy / Precision / Recall / F1）", lines=12)
+        cm_plot = gr.Plot(label="混淆矩陣")
+
+        btn_train.click(
+            fn=train_and_eval,
+            inputs=[state_df, split],
+            outputs=[state_model, report_box, cm_plot, train_msg]
+        )
+
+    with gr.Tab("Step 3｜測試/推論"):
+        gr.Markdown("單句測試或批次測試。")
+        with gr.Row():
+            test_text = gr.Textbox(label="單句輸入", placeholder="例：這家好吃到爆！", lines=2)
+            btn_pred = gr.Button("預測")
+        pred_out = gr.Textbox(label="結果", interactive=False)
+
+        btn_pred.click(fn=predict_one, inputs=[test_text, state_model], outputs=[pred_out])
+
+        gr.Markdown("---\n**批次測試**：在下方貼入多行文字（第一欄為 text），點擊預測。")
+        df_infer = gr.Dataframe(headers=["text"], row_count=5, col_count=1)
+        btn_batch = gr.Button("批次預測")
+        df_result = gr.Dataframe(label="批次結果（prob_pos=正面機率, pred=預測標籤）", interactive=False)
 
-    submit.click(fn=predict_one, inputs=inp, outputs=out)
+        btn_batch.click(fn=predict_batch, inputs=[df_infer, state_model], outputs=[df_result])
 
-# 在 Colab 若需要外網連結可改成 share=True
-demo.launch()  # 或 demo.launch(share=True)
+if __name__ == "__main__":
+    demo.launch()