Update app.py

app.py

@@ -4,204 +4,247 @@ import plotly.graph_objects as go
 import plotly.io as pio
 from zipfile import ZipFile
 from io import BytesIO
+import time
+import psutil
 
 # ================= 全局配置 =================
-FONT_NAME = "Noto Sans CJK TC"  # 精确匹配系统安装的字体名称
+FONT_NAME = "Noto Sans CJK TC"  # 根据构建日志验证实际字体名称
+MAX_STUDENTS = 50               # 最大处理学生数
+BATCH_SIZE = 5                  # 每批处理数量
+DELAY_SECONDS = 5               # 批处理间隔时间
 
 # 配置Kaleido（图片导出引擎）
 pio.kaleido.scope.default_format = "png"
 pio.kaleido.scope.default_width = 1200
 pio.kaleido.scope.default_height = 900
 pio.kaleido.scope.default_scale = 2
-pio.kaleido.scope.default_font = FONT_NAME  # 关键配置
+pio.kaleido.scope.default_font = FONT_NAME
 
 # ================= 核心函数 =================
 def load_data(uploaded_file):
-    """优化数据加载与验证"""
+    """优化内存使用的数据加载"""
     try:
-        df = pd.read_csv(uploaded_file, encoding='utf-8').dropna(how='all')
+        # 使用分块读取优化大文件处理
+        df = pd.read_csv(uploaded_file, 
+                        encoding='utf-8',
+                        dtype={'姓名': 'category'},
+                        usecols=lambda col: col in ['姓名', '平均', '總分', '國文', '英文', '數學', '自科', '社會'])
+        
+        # 清理无效数据
+        df = df.dropna(subset=['姓名']).head(MAX_STUDENTS)
         
-        # 数据验证
-        if '姓名' not in df.columns:
-            raise ValueError("CSV文件中缺少必要欄位：姓名")
-            
         # 自动识别数值列
-        numeric_columns = []
-        potential_numeric = ['平均', '總分', '國文', '英文', '數學', '自科', 
-                           '社會', '地理', '歷史', '公民', '物理', '化學', '生物']
-        for col in potential_numeric:
-            if col in df.columns:
-                try:
-                    df[col] = pd.to_numeric(df[col], errors='coerce')
-                    numeric_columns.append(col)
-                except:
-                    pass
-                    
-        if not numeric_columns:
-            raise ValueError("CSV文件中未找到有效的數值欄位")
+        numeric_cols = df.select_dtypes(include='number').columns.tolist()
+        if not numeric_cols:
+            raise ValueError("未找到有效的數值欄位")
             
-        return df, numeric_columns
+        return df, numeric_cols
     except Exception as e:
         st.error(f"數據加載錯誤：{str(e)}")
         return None, None
 
 def create_radar_chart(df, selected_rows, selected_columns, student_name=""):
-    """生成强化中文支持的雷达图"""
+    """高效生成雷达图"""
     colors = ['#1F77B4', '#FF7F0E', '#2CA02C', '#D62728', '#9467BD']
-    line_styles = ['solid', 'dot', 'dash', 'longdash', 'dashdot']
-
+    
     fig = go.Figure()
-
-    # 添加每个学生的雷达轨迹
-    for i, name in enumerate(selected_rows):
-        student_data = df[df['姓名'] == name][selected_columns].iloc[0]
-        fig.add_trace(go.Scatterpolar(
-            r=student_data.values,
-            theta=selected_columns,
-            fill='toself',
-            name=name,
-            line=dict(
-                color=colors[i % len(colors)],
-                dash=line_styles[i % len(line_styles)],
-                width=2
+    
+    # 使用生成器减少内存占用
+    def generate_traces():
+        for i, name in enumerate(selected_rows):
+            student_data = df[df['姓名'] == name][selected_columns].iloc[0]
+            yield go.Scatterpolar(
+                r=student_data.values,
+                theta=selected_columns,
+                fill='toself',
+                name=name,
+                line=dict(
+                    color=colors[i % len(colors)],
+                    width=2
+                )
             )
-        ))
-
-    # 动态调整最大值
-    max_value = df[selected_columns].max().max() * 1.1 if selected_columns else 100
-
-    # 统一字体配置
+    
+    fig.add_traces(list(generate_traces()))
+    
+    # 动态计算最大值
+    max_value = df[selected_columns].max().max() * 1.1
+    
+    # 优化布局配置
     fig.update_layout(
         polar=dict(
             radialaxis=dict(
-                visible=True,
                 range=[0, max_value],
-                tickfont=dict(size=14, family=FONT_NAME)
+                tickfont_size=14
             ),
             angularaxis=dict(
-                tickfont=dict(size=16, family=FONT_NAME),
-                rotation=20  # 调整标签旋转角度
+                tickfont_size=16,
+                rotation=20
             )
         ),
         showlegend=True,
         legend=dict(
-            font=dict(family=FONT_NAME, size=14),
-            orientation="v",
             x=1.2,
             y=0.5
         ),
-        title=dict(
-            text=f'{student_name} 成績比較圖' if student_name else '學生成績雷達圖',
-            font=dict(size=24, family=FONT_NAME),
-            x=0.5
-        ),
-        margin=dict(t=80, b=120, r=200),  # 增加右侧边距
-        font=dict(family=FONT_NAME, size=14)
+        margin=dict(t=50, b=100, r=200),
+        font_family=FONT_NAME,
+        title_font_size=24,
+        title_x=0.5
     )
+    
+    if student_name:
+        fig.update_layout(title_text=f"{student_name} 成績比較圖")
+    
     return fig
 
+# ================= 资源监控 =================
+def show_system_stats():
+    """显示系统资源状态"""
+    mem = psutil.virtual_memory()
+    stats = f"""
+    | 内存使用 | {mem.percent}% |
+    | CPU使用 | {psutil.cpu_percent()}% |
+    | 处理学生数 | {len(st.session_state.get('processed_students', []))} |
+    """
+    st.sidebar.markdown("### 系统状态")
+    st.sidebar.markdown(stats)
+
 # ================= 主界面逻辑 =================
 def main():
+    st.set_page_config(page_title="成績分析系統", layout="wide")
+    
+    # 初始化会话状态
+    if 'processed_students' not in st.session_state:
+        st.session_state.processed_students = []
+    
     st.title('學生成績雷達圖產生器')
-    st.markdown("""
-    <style>
-    .stDownloadButton>button { background-color: #4CAF50 !important; color: white !important; }
-    .stProgress > div > div > div { background-color: #4CAF50; }
-    </style>
-    """, unsafe_allow_html=True)
-
+    
+    # 文件上传区块
     uploaded_file = st.file_uploader("上傳CSV檔案", type=['csv'])
-
-    if uploaded_file is not None:
-        df, numeric_columns = load_data(uploaded_file)
+    
+    if uploaded_file:
+        df, numeric_cols = load_data(uploaded_file)
         
-        if df is not None and numeric_columns:
-            # ===== 科目选择区块 =====
-            st.write("### 選擇要比較的欄位")
+        if df is not None and numeric_cols:
+            # ===== 配置区块 =====
+            with st.sidebar:
+                st.header("分析設定")
+                
+                # 科目选择
+                use_preset = st.checkbox("使用預設科目", True)
+                preset_cols = ['平均', '國文', '英文', '數學', '自科', '社會']
+                available_preset = [c for c in preset_cols if c in numeric_cols]
+                selected_cols = st.multiselect(
+                    "選擇科目",
+                    options=numeric_cols,
+                    default=available_preset if use_preset else numeric_cols[:3]
+                )
+                
+                # 基准学生选择
+                baseline_students = st.multiselect(
+                    "選擇比較基準",
+                    options=df['姓名'].unique(),
+                    help="選擇要作為比較基準的學生"
+                )
             
-            # 快速选择预设科目
-            preset_columns = ['平均', '國文', '英文', '數學', '自科', '社會']
-            use_preset = st.checkbox("使用五科分數與平均比較")
-            available_preset = [col for col in preset_columns if col in numeric_columns]
+            # ===== 实时预览区块 =====
+            st.subheader("即時預覽")
+            preview_col1, preview_col2 = st.columns([3, 1])
             
-            selected_columns = st.multiselect(
-                '選擇科目',
-                options=numeric_columns,
-                default=available_preset if use_preset else numeric_columns[:5]
-            )
-
-            # ===== 学生选择区块 =====
-            st.write("### 選擇要比較的對象")
-            all_students = df['姓名'].tolist()
-            selected_students = st.multiselect('選擇學生', all_students)
-
-            # 实时显示雷达图
-            if selected_columns and selected_students:
-                try:
-                    fig = create_radar_chart(df, selected_students, selected_columns)
-                    st.plotly_chart(fig, use_container_width=True)
-                except Exception as e:
-                    st.error(f"生成雷達圖錯誤：{str(e)}")
-
-            # ===== 批量下载区块 =====
-            st.write("### 批次下載全班學生雷達圖")
-            baseline_students = st.multiselect("選擇比較基準", all_students)
-
-            if baseline_students and selected_columns:
-                progress_bar = st.progress(0)
-                image_data = {}
-                total = len([s for s in all_students if s not in baseline_students])
-
-                for idx, student in enumerate(all_students):
-                    if student in baseline_students:
-                        continue
-
+            with preview_col1:
+                selected_students = st.multiselect(
+                    "選擇預覽學生",
+                    options=df['姓名'].unique(),
+                    max_selections=3
+                )
+                
+                if selected_students and selected_cols:
                     try:
-                        # 生成每个学生的对比图
-                        fig = create_radar_chart(
-                            df, 
-                            baseline_students + [student],
-                            selected_columns,
-                            student_name=student
-                        )
-                        
-                        # 强制刷新字体配置
-                        fig.update_layout(font_family=FONT_NAME)
-                        
-                        # 生成图片字节数据
-                        img_bytes = pio.to_image(fig, format="png", engine="kaleido")
-                        image_data[student] = img_bytes
+                        fig = create_radar_chart(df, selected_students, selected_cols)
+                        st.plotly_chart(fig, use_container_width=True)
                     except Exception as e:
-                        st.error(f"生成 {student} 圖表失敗：{str(e)}")
+                        st.error(f"預覽生成失敗：{str(e)}")
+            
+            # ===== 批量处理区块 =====
+            st.divider()
+            st.subheader("批次輸出")
+            
+            if baseline_students and selected_cols:
+                # 初始化批处理
+                all_students = [s for s in df['姓名'] if s not in baseline_students]
+                total_batches = (len(all_students) + BATCH_SIZE - 1) // BATCH_SIZE
+                
+                # 进度控制
+                progress_bar = st.progress(0)
+                status_text = st.empty()
+                download_placeholder = st.empty()
+                
+                image_data = {}
+                
+                # 分批处理
+                for batch_num in range(total_batches):
+                    start_idx = batch_num * BATCH_SIZE
+                    end_idx = start_idx + BATCH_SIZE
+                    batch = all_students[start_idx:end_idx]
                     
-                    progress_bar.progress((idx+1)/len(all_students))
-
-                # 显示预览
-                if image_data:
-                    st.write("### 圖表預覽")
-                    cols = st.columns(3)
-                    for i, (name, img) in enumerate(image_data.items()):
-                        with cols[i % 3]:
-                            st.image(img, use_container_width=True)
-                            st.caption(f"{name} 比較圖")
-
-                    # 打包下载
-                    zip_buffer = BytesIO()
-                    with ZipFile(zip_buffer, "w") as zip_file:
-                        for name, data in image_data.items():
-                            zip_file.writestr(
-                                f"{name}_成績比較圖.png", 
-                                data
+                    # 生成当前批次
+                    for student in batch:
+                        try:
+                            start_time = time.time()
+                            
+                            fig = create_radar_chart(
+                                df, 
+                                baseline_students + [student],
+                                selected_cols,
+                                student_name=student
                             )
-                    zip_buffer.seek(0)
+                            
+                            # 优化图像生成
+                            img_bytes = pio.to_image(fig, format="png", engine="kaleido")
+                            image_data[student] = img_bytes
+                            st.session_state.processed_students.append(student)
+                            
+                            # 更新状态
+                            elapsed = time.time() - start_time
+                            status_text.info(f"""
+                            正在處理：{student}
+                            已用時間：{elapsed:.1f}秒
+                            剩餘批次：{total_batches - batch_num - 1}
+                            """)
+                            
+                        except Exception as e:
+                            st.error(f"學生 {student} 處理失敗：{str(e)}")
+                    
+                    # 更新进度
+                    progress = (batch_num + 1) / total_batches
+                    progress_bar.progress(progress)
                     
-                    st.download_button(
-                        label="⬇️ 下載全部圖表 (ZIP)",
-                        data=zip_buffer,
-                        file_name="學生成績比較圖.zip",
-                        mime="application/zip",
-                        use_container_width=True
-                    )
+                    # 添加延迟避免资源耗尽
+                    if batch_num < total_batches - 1:
+                        time.sleep(DELAY_SECONDS)
+                
+                # 完成处理
+                progress_bar.empty()
+                status_text.success(f"處理完成！共生成 {len(image_data)} 張圖表")
+                
+                # 打包下载
+                if image_data:
+                    with BytesIO() as zip_buffer:
+                        with ZipFile(zip_buffer, 'w') as zip_file:
+                            for name, data in image_data.items():
+                                zip_file.writestr(f"{name}.png", data)
+                        
+                        zip_buffer.seek(0)
+                        download_placeholder.download_button(
+                            "⬇️ 下載全部圖表",
+                            data=zip_buffer,
+                            file_name="scores_analysis.zip",
+                            mime="application/zip",
+                            use_container_width=True
+                        )
+            
+            # 显示系统状态
+            show_system_stats()
 
 if __name__ == "__main__":
     main()