AcidificationModel/api/app/routes.py

import sqlite3

from flask import Blueprint, request, jsonify,current_app
from werkzeug.security import generate_password_hash

from .model import predict, train_and_save_model, calculate_model_score
import pandas as pd
from . import db  # 从 app 包导入 db 实例
from sqlalchemy.engine.reflection import Inspector
from .database_models import Models, ModelParameters, Datasets, CurrentReduce, CurrentReflux
import os
from .utils import create_dynamic_table, allowed_file, infer_column_types, rename_columns_for_model_predict, \
    clean_column_names, rename_columns_for_model, insert_data_into_dynamic_table, insert_data_into_existing_table, \
    predict_to_Q, Q_to_t_ha, create_kriging
from sqlalchemy.orm import sessionmaker
import logging
from sqlalchemy import text, func
from .tasks import train_model_task
from datetime import datetime


# 配置日志
logging.basicConfig(level=logging.DEBUG)
logger = logging.getLogger(__name__)
# 创建蓝图 (Blueprint)，用于分离路由
bp = Blueprint('routes', __name__)

# 封装数据库连接函数
def get_db_connection():
    return sqlite3.connect('software_intro.db')

# 密码加密
def hash_password(password):
    return generate_password_hash(password)


def get_db():
    """ 获取数据库连接 """
    return sqlite3.connect(current_app.config['DATABASE'])


# 添加一个新的辅助函数来检查数据集大小并触发训练
def check_and_trigger_training(session, dataset_type, dataset_df):
    """
    检查当前数据集大小是否跨越新的阈值点并触发训练
    
    Args:
        session: 数据库会话
        dataset_type: 数据集类型 ('reduce' 或 'reflux')
        dataset_df: 数据集 DataFrame
        
    Returns:
        tuple: (是否触发训练, 任务ID)
    """
    try:
        # 根据数据集类型选择表
        table = CurrentReduce if dataset_type == 'reduce' else CurrentReflux
        
        # 获取当前记录数
        current_count = session.query(func.count()).select_from(table).scalar()
        
        # 获取新增的记录数(从request.files中获取的DataFrame长度)
        new_records = len(dataset_df)  # 需要从上层函数传入
        
        # 计算新增数据前的记录数
        previous_count = current_count - new_records
        
        # 设置阈值
        THRESHOLD = current_app.config['THRESHOLD']
        
        # 计算上一个阈值点(基于新增前的数据量)
        last_threshold = previous_count // THRESHOLD * THRESHOLD
        # 计算当前所在阈值点
        current_threshold = current_count // THRESHOLD * THRESHOLD
        
        # 检查是否跨越了新的阈值点
        if current_threshold > last_threshold and current_count >= THRESHOLD:
            # 触发异步训练任务
            task = train_model_task.delay(
                model_type=current_app.config['DEFAULT_MODEL_TYPE'],
                model_name=f'auto_trained_{dataset_type}_{current_threshold}',
                model_description=f'Auto trained model at {current_threshold} records threshold',
                data_type=dataset_type
            )
            return True, task.id
            
        return False, None
        
    except Exception as e:
        logging.error(f"检查并触发训练失败: {str(e)}")
        return False, None


@bp.route('/upload-dataset', methods=['POST'])
def upload_dataset():
    # 创建 session
    Session = sessionmaker(bind=db.engine)
    session = Session()
    
    try:
        if 'file' not in request.files:
            return jsonify({'error': 'No file part'}), 400
        file = request.files['file']
        if file.filename == '' or not allowed_file(file.filename):
            return jsonify({'error': 'No selected file or invalid file type'}), 400

        dataset_name = request.form.get('dataset_name')
        dataset_description = request.form.get('dataset_description', 'No description provided')
        dataset_type = request.form.get('dataset_type')
        if not dataset_type:
            return jsonify({'error': 'Dataset type is required'}), 400

        new_dataset = Datasets(
            Dataset_name=dataset_name,
            Dataset_description=dataset_description,
            Row_count=0,
            Status='Datasets_upgraded',
            Dataset_type=dataset_type,
            Uploaded_at=datetime.now()
        )
        session.add(new_dataset)
        session.commit()

        unique_filename = f"dataset_{new_dataset.Dataset_ID}.xlsx"
        upload_folder = current_app.config['UPLOAD_FOLDER']
        file_path = os.path.join(upload_folder, unique_filename)
        file.save(file_path)

        dataset_df = pd.read_excel(file_path)
        new_dataset.Row_count = len(dataset_df)
        new_dataset.Status = 'excel_file_saved success'
        session.commit()

        # 处理列名
        dataset_df = clean_column_names(dataset_df)
        dataset_df = rename_columns_for_model(dataset_df, dataset_type)

        column_types = infer_column_types(dataset_df)
        dynamic_table_class = create_dynamic_table(new_dataset.Dataset_ID, column_types)
        insert_data_into_dynamic_table(session, dataset_df, dynamic_table_class)

        # 根据 dataset_type 决定插入到哪个已有表
        if dataset_type == 'reduce':
            insert_data_into_existing_table(session, dataset_df, CurrentReduce)
        elif dataset_type == 'reflux':
            insert_data_into_existing_table(session, dataset_df, CurrentReflux)

        session.commit()

        # 在完成数据插入后，检查是否需要触发训练
        training_triggered, task_id = check_and_trigger_training(session, dataset_type, dataset_df)

        response_data = {
            'message': f'Dataset {dataset_name} uploaded successfully!',
            'dataset_id': new_dataset.Dataset_ID,
            'filename': unique_filename,
            'training_triggered': training_triggered
        }
        
        if training_triggered:
            response_data['task_id'] = task_id
            response_data['message'] += ' Auto-training has been triggered.'

        return jsonify(response_data), 201

    except Exception as e:
        session.rollback()
        logging.error('Failed to process the dataset upload:', exc_info=True)
        return jsonify({'error': str(e)}), 500
        
    finally:
        # 确保 session 总是被关闭
        if session:
            session.close()


@bp.route('/train-and-save-model', methods=['POST'])
def train_and_save_model_endpoint():
    # 创建 sessionmaker 实例
    Session = sessionmaker(bind=db.engine)
    session = Session()

    data = request.get_json()
    # 从请求中解析参数
    model_type = data.get('model_type')
    model_name = data.get('model_name')
    model_description = data.get('model_description')
    data_type = data.get('data_type')
    dataset_id = data.get('dataset_id', None)  # 默认为 None，如果未提供

    try:
        # 调用训练和保存模型的函数
        result = train_and_save_model(session, model_type, model_name, model_description, data_type, dataset_id)
        
        model_id = result[1] if result else None
        
        # 计算模型评分
        if model_id:
            model_info = session.query(Models).filter(Models.ModelID == model_id).first()
            if model_info:
                score = calculate_model_score(model_info)
                # 更新模型评分
                model_info.Performance_score = score
                session.commit()
                result = {'model_id': model_id, 'model_score': score}

        # 返回成功响应
        return jsonify({
            'message': 'Model trained and saved successfully', 
            'result': result
        }), 200

    except Exception as e:
        session.rollback()
        logging.error('Failed to process the model training:', exc_info=True)
        return jsonify({
            'error': 'Failed to train and save model', 
            'message': str(e)
        }), 500
    finally:
        session.close()


@bp.route('/predict', methods=['POST'])
def predict_route():
    # 创建 sessionmaker 实例
    Session = sessionmaker(bind=db.engine)
    session = Session()
    try:
        data = request.get_json()
        model_id = data.get('model_id')          # 提取模型名称
        parameters = data.get('parameters', {})  # 提取所有变量

        # 根据model_id获取模型Data_type
        model_info = session.query(Models).filter(Models.ModelID == model_id).first()
        if not model_info:
            return jsonify({'error': 'Model not found'}), 404
        data_type = model_info.Data_type

        input_data = pd.DataFrame([parameters])  # 转换参数为DataFrame
        # 如果为reduce，则不需要传入target_ph
        if data_type == 'reduce':
            # 获取传入的init_ph、target_ph参数
            init_ph = float(parameters.get('init_pH', 0.0))  # 默认值为0.0，防止None导致错误
            target_ph = float(parameters.get('target_pH', 0.0))  # 默认值为0.0，防止None导致错误

            # 从输入数据中删除'target_pH'列
            input_data = input_data.drop('target_pH', axis=1, errors='ignore')  # 使用errors='ignore'防止列不存在时出错

        input_data_rename = rename_columns_for_model_predict(input_data, data_type)  # 重命名列名以匹配模型字段
        predictions = predict(session, input_data_rename, model_id)  # 调用预测函数

        if data_type == 'reduce':
            predictions = predictions[0]
            # 将预测结果转换为Q
            Q = predict_to_Q(predictions, init_ph, target_ph)
            predictions = Q_to_t_ha(Q)  # 将Q转换为t/ha

            print(predictions)

        return jsonify({'result': predictions}), 200
    except Exception as e:
        logging.error('Failed to predict:', exc_info=True)
        return jsonify({'error': str(e)}), 400


# 为指定模型计算评分Performance_score，需要提供model_id
@bp.route('/score-model/<int:model_id>', methods=['POST'])
def score_model(model_id):
    # 创建 sessionmaker 实例
    Session = sessionmaker(bind=db.engine)
    session = Session()
    try:
        model_info = session.query(Models).filter(Models.ModelID == model_id).first()
        if not model_info:
            return jsonify({'error': 'Model not found'}), 404

        # 计算模型评分
        score = calculate_model_score(model_info)

        # 更新模型记录中的评分
        model_info.Performance_score = score
        session.commit()

        return jsonify({'message': 'Model scored successfully', 'score': score}), 200
    except Exception as e:
        logging.error('Failed to process the dataset upload:', exc_info=True)
        return jsonify({'error': str(e)}), 400
    finally:
        session.close()


@bp.route('/delete-dataset/<int:dataset_id>', methods=['DELETE'])
def delete_dataset_endpoint(dataset_id):
    """
    删除数据集的API接口
    
    @param dataset_id: 要删除的数据集ID
    @return: JSON响应
    """
    # 创建 sessionmaker 实例
    Session = sessionmaker(bind=db.engine)
    session = Session()
    try:
        # 查询数据集
        dataset = session.query(Datasets).filter_by(Dataset_ID=dataset_id).first()
        if not dataset:
            return jsonify({'error': '未找到数据集'}), 404

        # 检查是否有模型使用了该数据集
        models_using_dataset = session.query(Models).filter_by(DatasetID=dataset_id).all()
        if models_using_dataset:
            models_info = [{'ModelID': model.ModelID, 'Model_name': model.Model_name} for model in models_using_dataset]
            return jsonify({
                'error': '无法删除数据集,因为以下模型正在使用它',
                'models': models_info
            }), 400

        # 删除Excel文件
        filename = f"dataset_{dataset.Dataset_ID}.xlsx"
        file_path = os.path.join(current_app.config['UPLOAD_FOLDER'], filename)
        if os.path.exists(file_path):
            try:
                os.remove(file_path)
            except OSError as e:
                logger.error(f'删除文件失败: {str(e)}')
                return jsonify({'error': f'删除文件失败: {str(e)}'}), 500

        # 删除数据表
        table_name = f"dataset_{dataset.Dataset_ID}"
        session.execute(text(f"DROP TABLE IF EXISTS {table_name}"))

        # 删除数据集记录
        session.delete(dataset)
        session.commit()

        return jsonify({
            'message': '数据集删除成功',
            'deleted_files': [filename]
        }), 200

    except Exception as e:
        session.rollback()
        logger.error(f'删除数据集 {dataset_id} 失败:', exc_info=True)
        return jsonify({'error': str(e)}), 500
    finally:
        session.close()


@bp.route('/tables', methods=['GET'])
def list_tables():
    engine = db.engine  # 使用 db 实例的 engine
    inspector = Inspector.from_engine(engine)  # 创建 Inspector 对象
    table_names = inspector.get_table_names()  # 获取所有表名
    return jsonify(table_names)  # 以 JSON 形式返回表名列表


@bp.route('/models/<int:model_id>', methods=['GET'])
def get_model(model_id):
    """
    获取单个模型信息的API接口
    
    @param model_id: 模型ID
    @return: JSON响应
    """
    Session = sessionmaker(bind=db.engine)
    session = Session()
    
    try:
        model = session.query(Models).filter_by(ModelID=model_id).first()
        if model:
            return jsonify({
                'ModelID': model.ModelID,
                'Model_name': model.Model_name,
                'Model_type': model.Model_type,
                'Created_at': model.Created_at.strftime('%Y-%m-%d %H:%M:%S'),
                'Description': model.Description,
                'Performance_score': float(model.Performance_score) if model.Performance_score else None,
                'Data_type': model.Data_type
            })
        else:
            return jsonify({'message': '未找到模型'}), 404
            
    except Exception as e:
        logger.error(f'获取模型信息失败: {str(e)}')
        return jsonify({'error': '服务器内部错误', 'message': str(e)}), 500
        
    finally:
        session.close()


@bp.route('/model-parameters', methods=['GET'])
def get_all_model_parameters():
    """
    获取所有模型参数的API接口
    
    @return: JSON响应
    """
    Session = sessionmaker(bind=db.engine)
    session = Session()
    
    try:
        parameters = session.query(ModelParameters).all()
        if parameters:
            result = [
                {
                    'ParamID': param.ParamID,
                    'ModelID': param.ModelID,
                    'ParamName': param.ParamName,
                    'ParamValue': param.ParamValue
                }
                for param in parameters
            ]
            return jsonify(result)
        else:
            return jsonify({'message': '未找到任何参数'}), 404
            
    except Exception as e:
        logger.error(f'获取所有模型参数失败: {str(e)}')
        return jsonify({'error': '服务器内部错误', 'message': str(e)}), 500
        
    finally:
        session.close()


@bp.route('/models/<int:model_id>/parameters', methods=['GET'])
def get_model_parameters(model_id):
    try:
        model = Models.query.filter_by(ModelID=model_id).first()
        if model:
            # 获取该模型的所有参数
            parameters = [
                {
                    'ParamID': param.ParamID,
                    'ParamName': param.ParamName,
                    'ParamValue': param.ParamValue
                }
                for param in model.parameters
            ]
            
            # 返回模型参数信息
            return jsonify({
                'ModelID': model.ModelID,
                'ModelName': model.ModelName,
                'ModelType': model.ModelType,
                'CreatedAt': model.CreatedAt.strftime('%Y-%m-%d %H:%M:%S'),
                'Description': model.Description,
                'Parameters': parameters
            })
        else:
            return jsonify({'message': 'Model not found'}), 404
    except Exception as e:
        return jsonify({'error': 'Internal server error', 'message': str(e)}), 500


@bp.route('/train-model-async', methods=['POST'])
def train_model_async():
    """
    异步训练模型的API接口
    """
    try:
        data = request.get_json()
        
        # 从请求中获取参数
        model_type = data.get('model_type')
        model_name = data.get('model_name')
        model_description = data.get('model_description')
        data_type = data.get('data_type')
        dataset_id = data.get('dataset_id', None)
        
        # 验证必要参数
        if not all([model_type, model_name, data_type]):
            return jsonify({
                'error': 'Missing required parameters'
            }), 400
            
        # 如果提供了dataset_id，验证数据集是否存在
        if dataset_id:
            Session = sessionmaker(bind=db.engine)
            session = Session()
            try:
                dataset = session.query(Datasets).filter_by(Dataset_ID=dataset_id).first()
                if not dataset:
                    return jsonify({
                        'error': f'Dataset with ID {dataset_id} not found'
                    }), 404
            finally:
                session.close()
            
        # 启动异步任务
        task = train_model_task.delay(
            model_type=model_type,
            model_name=model_name,
            model_description=model_description,
            data_type=data_type,
            dataset_id=dataset_id
        )
        
        # 返回任务ID
        return jsonify({
            'task_id': task.id,
            'message': 'Model training started'
        }), 202
        
    except Exception as e:
        logging.error('Failed to start async training task:', exc_info=True)
        return jsonify({
            'error': str(e)
        }), 500


@bp.route('/task-status/<task_id>', methods=['GET'])
def get_task_status(task_id):
    """
    获取异步任务状态的API接口
    """
    try:
        task = train_model_task.AsyncResult(task_id)
        
        if task.state == 'PENDING':
            response = {
                'state': task.state,
                'status': 'Task is waiting for execution'
            }
        elif task.state == 'FAILURE':
            response = {
                'state': task.state,
                'status': 'Task failed',
                'error': task.info.get('error') if isinstance(task.info, dict) else str(task.info)
            }
        elif task.state == 'SUCCESS':
            response = {
                'state': task.state,
                'status': 'Task completed successfully',
                'result': task.get()
            }
        else:
            response = {
                'state': task.state,
                'status': 'Task is in progress'
            }
            
        return jsonify(response), 200
        
    except Exception as e:
        return jsonify({
            'error': str(e)
        }), 500


@bp.route('/delete-model/<int:model_id>', methods=['DELETE'])
def delete_model_route(model_id):
    # 将URL参数转换为布尔值
    delete_dataset_param = request.args.get('delete_dataset', 'False').lower() == 'true'
    
    # 调用原始函数
    return delete_model(model_id, delete_dataset=delete_dataset_param)

def delete_model(model_id, delete_dataset=False):
    """
    删除指定模型的API接口
    
    @param model_id: 要删除的模型ID
    @query_param delete_dataset: 布尔值，是否同时删除关联的数据集，默认为False
    @return: JSON响应
    """
    Session = sessionmaker(bind=db.engine)
    session = Session()
    
    try:
        # 查询模型信息
        model = session.query(Models).filter_by(ModelID=model_id).first()
        if not model:
            return jsonify({'error': '未找到指定模型'}), 404
        
        dataset_id = model.DatasetID
        
        # 1. 先删除模型记录
        session.delete(model)
        session.commit()
        
        # 2. 删除模型文件
        model_file = f"rf_model_{model_id}.pkl"
        model_path = os.path.join(current_app.config['MODEL_SAVE_PATH'], model_file)
        if os.path.exists(model_path):
            try:
                os.remove(model_path)
            except OSError as e:
                # 如果删除文件失败,回滚数据库操作
                session.rollback()
                logger.error(f'删除模型文件失败: {str(e)}')
                return jsonify({'error': f'删除模型文件失败: {str(e)}'}), 500

        # 3. 如果需要删除关联的数据集
        if delete_dataset and dataset_id:
            try:
                dataset_response = delete_dataset_endpoint(dataset_id)
                if not isinstance(dataset_response, tuple) or dataset_response[1] != 200:
                    # 如果删除数据集失败,回滚之前的操作
                    session.rollback()
                    return jsonify({
                        'error': '删除关联数据集失败',
                        'dataset_error': dataset_response[0].get_json() if hasattr(dataset_response[0], 'get_json') else str(dataset_response[0])
                    }), 500
            except Exception as e:
                session.rollback()
                logger.error(f'删除关联数据集失败: {str(e)}')
                return jsonify({'error': f'删除关联数据集失败: {str(e)}'}), 500

        response_data = {
            'message': '模型删除成功',
            'deleted_files': [model_file]
        }
        
        if delete_dataset:
            response_data['dataset_info'] = {
                'dataset_id': dataset_id,
                'message': '关联数据集已删除'
            }
        return jsonify(response_data), 200
        
    except Exception as e:
        session.rollback()
        logger.error(f'删除模型 {model_id} 失败:', exc_info=True)
        return jsonify({'error': str(e)}), 500
    finally:
        session.close()


# 添加一个新的API端点来清空指定数据集
@bp.route('/clear-dataset/<string:data_type>', methods=['DELETE'])
def clear_dataset(data_type):
    """
    清空指定类型的数据集并递增计数

    @param data_type: 数据集类型 ('reduce' 或 'reflux')
    @return: JSON响应
    """
    # 创建 sessionmaker 实例
    Session = sessionmaker(bind=db.engine)
    session = Session()
    
    try:
        # 根据数据集类型选择表
        if data_type == 'reduce':
            table = CurrentReduce
            table_name = 'current_reduce'
        elif data_type == 'reflux':
            table = CurrentReflux
            table_name = 'current_reflux'
        else:
            return jsonify({'error': '无效的数据集类型'}), 400
        
        # 清空表内容
        session.query(table).delete()
        
        # 重置自增主键计数器
        session.execute(text(f"DELETE FROM sqlite_sequence WHERE name='{table_name}'"))
                
        session.commit()
        
        return jsonify({'message': f'{data_type} 数据集已清空并重置计数器'}), 200
    
    except Exception as e:
        session.rollback()
        return jsonify({'error': str(e)}), 500
    
    finally:
        session.close()


@bp.route('/update-threshold', methods=['POST'])
def update_threshold():
    """
    更新训练阈值的API接口
    
    @body_param threshold: 新的阈值值(整数)
    @return: JSON响应
    """
    try:
        data = request.get_json()
        new_threshold = data.get('threshold')
        
        # 验证新阈值
        if not isinstance(new_threshold, (int, float)) or new_threshold <= 0:
            return jsonify({
                'error': '无效的阈值值，必须为正数'
            }), 400
            
        # 更新当前应用的阈值配置
        current_app.config['THRESHOLD'] = int(new_threshold)
        
        return jsonify({
            'success': True,
            'message': f'阈值已更新为 {new_threshold}',
            'new_threshold': new_threshold
        })
        
    except Exception as e:
        logging.error(f"更新阈值失败: {str(e)}")
        return jsonify({
            'error': f'更新阈值失败: {str(e)}'
        }), 500


@bp.route('/get-threshold', methods=['GET'])
def get_threshold():
    """
    获取当前训练阈值的API接口
    
    @return: JSON响应
    """
    try:
        current_threshold = current_app.config['THRESHOLD']
        default_threshold = current_app.config['DEFAULT_THRESHOLD']
        
        return jsonify({
            'current_threshold': current_threshold,
            'default_threshold': default_threshold
        })
        
    except Exception as e:
        logging.error(f"获取阈值失败: {str(e)}")
        return jsonify({
            'error': f'获取阈值失败: {str(e)}'
        }), 500

@bp.route('/set-current-dataset/<string:data_type>/<int:dataset_id>', methods=['POST'])
def set_current_dataset(data_type, dataset_id):
    """
    将指定数据集设置为current数据集
    
    @param data_type: 数据集类型 ('reduce' 或 'reflux')
    @param dataset_id: 要设置为current的数据集ID
    @return: JSON响应
    """
    Session = sessionmaker(bind=db.engine)
    session = Session()
    
    try:
        # 验证数据集存在且类型匹配
        dataset = session.query(Datasets)\
            .filter_by(Dataset_ID=dataset_id, Dataset_type=data_type)\
            .first()
            
        if not dataset:
            return jsonify({
                'error': f'未找到ID为 {dataset_id} 且类型为 {data_type} 的数据集'
            }), 404
            
        # 根据数据类型选择表
        if data_type == 'reduce':
            table = CurrentReduce
            table_name = 'current_reduce'
        elif data_type == 'reflux':
            table = CurrentReflux
            table_name = 'current_reflux'
        else:
            return jsonify({'error': '无效的数据集类型'}), 400
            
        # 清空current表
        session.query(table).delete()
        
        # 重置自增主键计数器
        session.execute(text(f"DELETE FROM sqlite_sequence WHERE name='{table_name}'"))
        
        # 从指定数据集复制数据到current表
        dataset_table_name = f"dataset_{dataset_id}"
        copy_sql = text(f"INSERT INTO {table_name} SELECT * FROM {dataset_table_name}")
        session.execute(copy_sql)
        
        session.commit()
        
        return jsonify({
            'message': f'{data_type} current数据集已设置为数据集 ID: {dataset_id}',
            'dataset_id': dataset_id,
            'dataset_name': dataset.Dataset_name,
            'row_count': dataset.Row_count
        }), 200
        
    except Exception as e:
        session.rollback()
        logger.error(f'设置current数据集失败: {str(e)}')
        return jsonify({'error': str(e)}), 500
        
    finally:
        session.close()

@bp.route('/get-model-history/<string:data_type>', methods=['GET'])
def get_model_history(data_type):
    """
    获取模型训练历史数据的API接口
    
    @param data_type: 数据集类型 ('reduce' 或 'reflux')
    @return: JSON响应，包含时间序列的模型性能数据
    """
    Session = sessionmaker(bind=db.engine)
    session = Session()
    
    try:
        # 查询所有自动生成的数据集，按时间排序
        datasets = session.query(Datasets).filter(
            Datasets.Dataset_type == data_type,
            Datasets.Dataset_description == f"Automatically generated dataset for type {data_type}"
        ).order_by(Datasets.Uploaded_at).all()
        
        history_data = []
        for dataset in datasets:
            # 查找对应的自动训练模型
            model = session.query(Models).filter(
                Models.DatasetID == dataset.Dataset_ID,
                Models.Model_name.like(f'auto_trained_{data_type}_%')
            ).first()
            
            if model and model.Performance_score is not None:
                # 直接使用数据库中的时间，不进行格式化（保持与created_at相同的时区）
                created_at = model.Created_at.isoformat() if model.Created_at else None
                
                history_data.append({
                    'dataset_id': dataset.Dataset_ID,
                    'row_count': dataset.Row_count,
                    'model_id': model.ModelID,
                    'model_name': model.Model_name,
                    'performance_score': float(model.Performance_score),
                    'timestamp': created_at
                })
        
        # 按时间戳排序
        history_data.sort(key=lambda x: x['timestamp'] if x['timestamp'] else '')
        
        # 构建返回数据，分离各个指标序列便于前端绘图
        response_data = {
            'data_type': data_type,
            'timestamps': [item['timestamp'] for item in history_data],
            'row_counts': [item['row_count'] for item in history_data],
            'performance_scores': [item['performance_score'] for item in history_data],
            'model_details': history_data  # 保留完整数据供前端使用
        }
        
        return jsonify(response_data), 200
        
    except Exception as e:
        logger.error(f'获取模型历史数据失败: {str(e)}', exc_info=True)
        return jsonify({'error': str(e)}), 500
        
    finally:
        session.close()

@bp.route('/batch-delete-datasets', methods=['POST'])
def batch_delete_datasets():
    """
    批量删除数据集的API接口
    
    @body_param dataset_ids: 要删除的数据集ID列表
    @return: JSON响应
    """
    try:
        data = request.get_json()
        dataset_ids = data.get('dataset_ids', [])
        
        if not dataset_ids:
            return jsonify({'error': '未提供数据集ID列表'}), 400
            
        results = {
            'success': [],
            'failed': [],
            'protected': []  # 被模型使用的数据集
        }
        
        for dataset_id in dataset_ids:
            try:
                # 调用单个删除接口
                response = delete_dataset_endpoint(dataset_id)
                
                # 解析响应
                if response[1] == 200:
                    results['success'].append(dataset_id)
                elif response[1] == 400 and 'models' in response[0].json:
                    # 数据集被模型保护
                    results['protected'].append({
                        'id': dataset_id,
                        'models': response[0].json['models']
                    })
                else:
                    results['failed'].append({
                        'id': dataset_id,
                        'reason': response[0].json.get('error', '删除失败')
                    })
                    
            except Exception as e:
                logger.error(f'删除数据集 {dataset_id} 失败: {str(e)}')
                results['failed'].append({
                    'id': dataset_id,
                    'reason': str(e)
                })
        
        # 构建响应消息
        message = f"成功删除 {len(results['success'])} 个数据集"
        if results['protected']:
            message += f", {len(results['protected'])} 个数据集被保护"
        if results['failed']:
            message += f", {len(results['failed'])} 个数据集删除失败"
            
        return jsonify({
            'message': message,
            'results': results
        }), 200
        
    except Exception as e:
        logger.error(f'批量删除数据集失败: {str(e)}')
        return jsonify({'error': str(e)}), 500

@bp.route('/batch-delete-models', methods=['POST'])
def batch_delete_models():
    """
    批量删除模型的API接口
    
    @body_param model_ids: 要删除的模型ID列表
    @query_param delete_datasets: 布尔值，是否同时删除关联的数据集，默认为False
    @return: JSON响应
    """
    try:
        data = request.get_json()
        model_ids = data.get('model_ids', [])
        delete_datasets = request.args.get('delete_datasets', 'false').lower() == 'true'
        
        if not model_ids:
            return jsonify({'error': '未提供模型ID列表'}), 400
            
        results = {
            'success': [],
            'failed': [],
            'datasets_deleted': []  # 如果delete_datasets为true，记录被删除的数据集
        }
        
        for model_id in model_ids:
            try:
                # 调用单个删除接口
                response = delete_model(model_id, delete_dataset=delete_datasets)
                
                # 解析响应
                if response[1] == 200:
                    results['success'].append(model_id)
                    # 如果删除了关联数据集，记录数据集ID
                    if 'dataset_info' in response[0].json:
                        results['datasets_deleted'].append(
                            response[0].json['dataset_info']['dataset_id']
                        )
                else:
                    results['failed'].append({
                        'id': model_id,
                        'reason': response[0].json.get('error', '删除失败')
                    })
                    
            except Exception as e:
                logger.error(f'删除模型 {model_id} 失败: {str(e)}')
                results['failed'].append({
                    'id': model_id,
                    'reason': str(e)
                })
        
        # 构建响应消息
        message = f"成功删除 {len(results['success'])} 个模型"
        if results['datasets_deleted']:
            message += f", {len(results['datasets_deleted'])} 个关联数据集"
        if results['failed']:
            message += f", {len(results['failed'])} 个模型删除失败"
            
        return jsonify({
            'message': message,
            'results': results
        }), 200
        
    except Exception as e:
        logger.error(f'批量删除模型失败: {str(e)}')
        return jsonify({'error': str(e)}), 500

@bp.route('/kriging_interpolation', methods=['POST'])
def kriging_interpolation():
    try:
        data = request.get_json()
        required = ['file_name', 'emission_column', 'points']
        if not all(k in data for k in required):
            return jsonify({"error": "Missing parameters"}), 400

        # 添加坐标顺序验证
        points = data['points']
        if not all(len(pt) == 2 and isinstance(pt[0], (int, float)) for pt in points):
            return jsonify({"error": "Invalid points format"}), 400

        result = create_kriging(
            data['file_name'],
            data['emission_column'],
            data['points']
        )
        return jsonify(result)
    except Exception as e:
        return jsonify({"error": str(e)}), 500

# 显示切换模型
@bp.route('/models', methods=['GET'])
def get_models():
    session = None
    try:
        # 创建 session
        Session = sessionmaker(bind=db.engine)
        session = Session()

        # 查询所有模型
        models = session.query(Models).all()

        logger.debug(f"Models found: {models}")  # 打印查询的模型数据

        if not models:
            return jsonify({'message': 'No models found'}), 404

        # 将模型数据转换为字典列表
        models_list = [
            {
                'ModelID': model.ModelID,
                'ModelName': model.Model_name,
                'ModelType': model.Model_type,
                'CreatedAt': model.Created_at.strftime('%Y-%m-%d %H:%M:%S'),
                'Description': model.Description,
                'DatasetID': model.DatasetID,
                'ModelFilePath': model.ModelFilePath,
                'DataType': model.Data_type,
                'PerformanceScore': model.Performance_score
            }
            for model in models
        ]

        return jsonify(models_list), 200

    except Exception as e:
        return jsonify({'error': str(e)}), 400
    finally:
        if session:
            session.close()