AI 驱动的数据库 TDSQL-C 操作与电商可视分析
qq_45424503 2024-10-25 12:01:01 阅读 72
TDSQL-C介绍
TDSQL-C是腾讯云自研的新一代高性能、高可用、高扩展性的企业级分布式云数据库，该数据库融合了传统数据库、云计算与新硬件技术的优势，为企业提供了全面且强大的数据管理解决方案。
TDSQL-C采用微服务架构，计算与存储分离，日志即数据库。存储构建在腾讯云分布式云存储服务之上，计算层全面兼容开源数据库引擎。通过基于SPDK和RDMA的零拷贝技术，减少操作系统上下文切换和数据拷贝引起的性能损耗。
整体架构流程
1.大语言模型根据自然语言生成SQL语句
2.使用SQL语句调用 TDSQL-C 数据库中保存的数据
3.把调用出来的数据发送给大语言模型
4.大语言模型根据数据生成可视化的python图表代码
5.运行Python代码产出可视化的数据图像
概述总结
大语言模型调用数据库可以灵活方便地调用
大语言模型回答结果不唯一，调用的结果具有一定的随机性，使得存在调用失败与出现错误的可能
可能的解决方案：
同时调用多个大模型，多次调用大模型，将返回的结果进行逻辑判断之后再选择
标题以下是生成图表的提示词
“”"
以下提供当前python环境已经安装的pip包集合:
{installed_packages};
请根据data提供的信息，生成是一个适合展示数据的plotly的图表的可执行代码,要求如下：
1.不要导入没有安装的pip包代码
2.如果存在多个数据类别，尽量使用柱状图，循环生成时图表中对不同数据请使用不同颜色区分，
3.图表要生成图片格式,保存在当前文件夹下即可，名称固定为：图表.png,
4.我需要您生成的代码是没有 Markdown 标记的，纯粹的编程语言代码。
5.生成的代码请注意将所有依赖包提前导入，
6.不要使用iplot等需要特定环境的代码
7.请注意数据之间是否可以转换，使用正确的代码
8.不需要生成注释
data:{data}
这是查询的sql语句与文本:
sql:{sql}
question:{question}
返回数据要求:
仅仅返回python代码，不要有额外的字符
“”"
配置TDSQL-C（数据库）
https://cloud.tencent.com/product/tdsqlc
配置数据库的过程中可能需要先配置子网
新建数据库
导入数据
<code>CREATE TABLE `ecommerce_sales_stats` (
`category_id` int NOT NULL COMMENT '分类ID（主键）',
`category_name` varchar(100) NOT NULL COMMENT '分类名称',
`total_sales` decimal(15,2) NOT NULL COMMENT '总销售额',
`steam_sales` decimal(15,2) NOT NULL COMMENT 'Steam平台销售额',
`offline_sales` decimal(15,2) NOT NULL COMMENT '线下实体销售额',
`official_online_sales` decimal(15,2) NOT NULL COMMENT '官方在线销售额',
PRIMARY KEY (`category_id`)
) ENGINE=INNODB DEFAULT CHARSET=utf8mb4 AUTO_INCREMENT=1 COMMENT='电商分类销售统计表';code>
INSERT INTO `ecommerce_sales_stats` VALUES (1,'电子产品',150000.00,80000.00,30000.00,40000.00),(2,'服装',120000.00,20000.00,60000.00,40000.00),(3,'家居用品',90000.00,10000.00,50000.00,30000.00),(4,'玩具',60000.00,5000.00,30000.00,25000.00),(5,'书籍',45000.00,2000.00,20000.00,23000.00),(6,'运动器材',70000.00,15000.00,25000.00,30000.00),(7,'美容护肤',80000.00,10000.00,30000.00,40000.00),(8,'食品',50000.00,5000.00,25000.00,20000.00),(9,'珠宝首饰',30000.00,2000.00,10000.00,18000.00),(10,'汽车配件',40000.00,10000.00,15000.00,25000.00),(11,'手机配件',75000.00,30000.00,20000.00,25000.00),(12,'电脑配件',85000.00,50000.00,15000.00,20000.00),(13,'摄影器材',50000.00,20000.00,15000.00,15000.00),(14,'家电',120000.00,60000.00,30000.00,30000.00),(15,'宠物用品',30000.00,3000.00,12000.00,16800.00),(16,'母婴用品',70000.00,10000.00,30000.00,30000.00),(17,'旅行用品',40000.00,5000.00,15000.00,20000.00),(18,'艺术品',25000.00,1000.00,10000.00,14000.00),(19,'健康产品',60000.00,8000.00,25000.00,27000.00),(20,'办公用品',55000.00,2000.00,20000.00,33000.00);
CREATE TABLE `users` (
`user_id` int NOT NULL AUTO_INCREMENT COMMENT '用户ID（主键，自增）',
`full_name` varchar(100) NOT NULL COMMENT '用户全名',
`username` varchar(50) NOT NULL COMMENT '用户名',
`email` varchar(100) NOT NULL COMMENT '用户邮箱',
`password_hash` varchar(255) NOT NULL COMMENT '用户密码的哈希值',
`created_at` datetime DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
`updated_at` datetime DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
`is_active` tinyint(1) DEFAULT '1' COMMENT '是否激活',
PRIMARY KEY (`user_id`),
UNIQUE KEY `email` (`email`)
) ENGINE=INNODB AUTO_INCREMENT=1 DEFAULT CHARSET=utf8mb4 COMMENT='用户表';code>
INSERT INTO `users` VALUES (1,'张伟','zhangwei','zhangwei@example.com','hashed_password_1','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(2,'李娜','lina','lina@example.com','hashed_password_2','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(3,'王芳','wangfang','wangfang@example.com','hashed_password_3','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(4,'刘洋','liuyang','liuyang@example.com','hashed_password_4','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(5,'陈杰','chenjie','chenjie@example.com','hashed_password_5','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(6,'杨静','yangjing','yangjing@example.com','hashed_password_6','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(7,'赵强','zhaoqiang','zhaoqiang@example.com','hashed_password_7','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(8,'黄丽','huangli','huangli@example.com','hashed_password_8','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(9,'周杰','zhoujie','zhoujie@example.com','hashed_password_9','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(10,'吴敏','wumin','wumin@example.com','hashed_password_10','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(11,'郑伟','zhengwei','zhengwei@example.com','hashed_password_11','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(12,'冯婷','fengting','fengting@example.com','hashed_password_12','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(13,'蔡明','caiming','caiming@example.com','hashed_password_13','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(14,'潘雪','panxue','panxue@example.com','hashed_password_14','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(15,'蒋磊','jianglei','jianglei@example.com','hashed_password_15','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(16,'陆佳','lujia','lujia@example.com','hashed_password_16','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(17,'邓超','dengchao','dengchao@example.com','hashed_password_17','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(18,'任丽','renli','renli@example.com','hashed_password_18','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(19,'彭涛','pengtao','pengtao@example.com','hashed_password_19','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(20,'方圆','fangyuan','fangyuan@example.com','hashed_password_20','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(21,'段飞','duanfei','duanfei@example.com','hashed_password_21','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(22,'雷鸣','leiming','leiming@example.com','hashed_password_22','2024-08-18 04:07:18','2024-08-18 04:07:18',1),(23,'贾玲','jialing','jialing@example.com','hashed_password_23','2024-08-18 04:07:18','2024-08-18 04:07:18',1);
CREATE TABLE `orders` (
`order_id` int NOT NULL AUTO_INCREMENT,
`user_id` int DEFAULT NULL,
`order_amount` decimal(10,2) DEFAULT NULL,
`order_status` varchar(20) DEFAULT NULL,
`order_time` datetime DEFAULT NULL,
PRIMARY KEY (`order_id`)
) ENGINE=InnoDB AUTO_INCREMENT=1 DEFAULT CHARSET=utf8mb4 ;
INSERT INTO `orders` VALUES (1,3,150.50,'已支付','2024-08-23 10:01:00'),(2,7,89.20,'待支付','2024-08-23 10:03:15'),(3,12,230.00,'已支付','2024-08-23 10:05:30'),(4,2,99.90,'已发货','2024-08-23 10:07:45'),(5,15,120.00,'待发货','2024-08-23 10:10:00'),(6,21,180.50,'已支付','2024-08-23 10:12:15'),(7,4,105.80,'待支付','2024-08-23 10:14:30'),(8,18,210.00,'已支付','2024-08-23 10:16:45'),(9,6,135.20,'已发货','2024-08-23 10:19:00'),(10,10,160.00,'待发货','2024-08-23 10:21:15'),(11,1,110.50,'已支付','2024-08-23 10:23:30'),(12,22,170.80,'待支付','2024-08-23 10:25:45'),(13,8,145.20,'已发货','2024-08-23 10:28:00'),(14,16,190.00,'待发货','2024-08-23 10:30:15'),(15,11,125.50,'已支付','2024-08-23 10:32:30'),(16,19,165.20,'待支付','2024-08-23 10:34:45'),(17,5,130.00,'已发货','2024-08-23 10:37:00'),(18,20,175.80,'待发货','2024-08-23 10:39:15'),(19,13,140.50,'已支付','2024-08-23 10:41:30'),(20,14,155.20,'待支付','2024-08-23 10:43:45'),(21,9,135.50,'已发货','2024-08-23 10:46:00'),(22,23,185.80,'待发货','2024-08-23 10:48:15'),(23,17,160.50,'已支付','2024-08-23 10:50:30'),(24,12,145.20,'待支付','2024-08-23 10:52:45'),(25,3,130.00,'已发货','2024-08-23 10:55:00'),(26,8,115.50,'已支付','2024-08-23 10:57:15'),(27,19,120.20,'待支付','2024-08-23 10:59:30'),(28,6,145.50,'已发货','2024-08-23 11:01:45'),(29,14,130.20,'待支付','2024-08-23 11:04:00'),(30,5,125.50,'已支付','2024-08-23 11:06:15'),(31,21,135.20,'待支付','2024-08-23 11:08:30'),(32,7,140.50,'已发货','2024-08-23 11:10:45'),(33,16,120.20,'待支付','2024-08-23 11:13:00'),(34,10,135.50,'已支付','2024-08-23 11:15:15'),(35,2,140.20,'待支付','2024-08-23 11:17:30'),(36,12,145.20,'待支付','2024-08-23 12:00:00'),(37,15,130.20,'已支付','2024-08-23 12:02:15'),(38,20,125.50,'待发货','2024-08-23 12:04:30'),(39,17,135.20,'已支付','2024-08-23 12:06:45'),(40,4,140.50,'待支付','2024-08-23 12:09:00'),(41,10,120.20,'已发货','2024-08-23 12:11:15'),(42,13,135.50,'已支付','2024-08-23 12:13:30'),(43,18,145.20,'待支付','2024-08-23 12:15:45'),(44,6,130.20,'已发货','2024-08-23 12:18:00'),(45,11,125.50,'已支付','2024-08-23 12:20:15'),(46,19,135.20,'待支付','2024-08-23 12:22:30'),(47,5,140.50,'已发货','2024-08-23 12:24:45'),(48,20,120.20,'待支付','2024-08-23 12:27:00'),(49,17,135.50,'已支付','2024-08-23 12:29:15'),(50,4,145.20,'待支付','2024-08-23 12:31:30'),(51,10,130.20,'已发货','2024-08-23 12:33:45'),(52,13,125.50,'已支付','2024-08-23 12:36:00'),(53,18,135.20,'待支付','2024-08-23 12:38:15'),(54,6,140.50,'已发货','2024-08-23 12:40:30'),(55,11,120.20,'待支付','2024-08-23 12:42:45'),(56,19,135.50,'已支付','2024-08-23 12:45:00'),(57,5,145.20,'待支付','2024-08-23 12:47:15'),(58,20,130.20,'已发货','2024-08-23 12:49:30'),(59,17,125.50,'已支付','2024-08-23 13:01:45'),(60,4,135.20,'待支付','2024-08-23 13:04:00'),(61,10,140.50,'已发货','2024-08-23 13:06:15'),(62,13,120.20,'待支付','2024-08-23 13:08:30'),(63,18,135.50,'已支付','2024-08-23 13:10:45'),(64,6,145.20,'待支付','2024-08-23 13:13:00'),(65,11,130.20,'已发货','2024-08-23 13:15:15'),(66,19,125.50,'已支付','2024-08-23 13:17:30'),(67,5,135.20,'待支付','2024-08-23 13:19:45'),(68,20,140.50,'已发货','2024-08-23 13:22:00'),(69,17,120.20,'待支付','2024-08-23 13:24:15'),(70,4,135.50,'已支付','2024-08-23 13:26:30'),(71,10,145.20,'待支付','2024-08-23 13:28:45'),(72,13,130.20,'已发货','2024-08-23 13:31:00'),(73,18,125.50,'已支付','2024-08-23 13:33:15'),(74,6,135.20,'待支付','2024-08-23 13:35:30'),(75,11,140.50,'已发货','2024-08-23 13:37:45'),(76,19,120.20,'待支付','2024-08-23 13:40:00'),(77,5,135.50,'已支付','2024-08-23 13:42:15'),(78,20,145.20,'待支付','2024-08-23 13:44:30'),(79,17,130.20,'已发货','2024-08-23 13:46:45'),(80,4,125.50,'已支付','2024-08-23 13:49:00'),(81,10,135.20,'待支付','2024-08-23 13:51:15'),(82,13,140.50,'已发货','2024-08-23 13:53:30'),(83,18,120.20,'待支付','2024-08-23 13:55:45'),(84,6,135.50,'已支付','2024-08-23 13:58:00'),(85,11,145.20,'待支付','2024-08-23 14:00:15'),(86,19,130.20,'已发货','2024-08-23 14:02:30'),(87,5,125.50,'已支付','2024-08-23 14:04:45'),(88,20,135.20,'待支付','2024-08-23 14:07:00'),(89,17,140.50,'已发货','2024-08-23 14:09:15'),(90,4,120.20,'待支付','2024-08-23 14:11:30'),(91,10,135.50,'已支付','2024-08-23 14:13:45'),(92,13,145.20,'待支付','2024-08-23 14:16:00'),(93,18,130.20,'已发货','2024-08-23 14:18:15'),(94,6,125.50,'已支付','2024-08-23 14:20:30'),(95,11,135.20,'待支付','2024-08-23 14:22:45'),(96,19,140.50,'已发货','2024-08-23 14:25:00'),(97,5,120.20,'待支付','2024-08-23 14:27:15'),(98,20,135.50,'已支付','2024-08-23 14:29:30'),(99,17,145.20,'待支付','2024-08-23 14:31:45'),(100,4,130.20,'已发货','2024-08-23 14:34:00'),(101,10,125.50,'已支付','2024-08-23 14:36:15'),(102,13,135.20,'待支付','2024-08-23 14:38:30'),(103,18,140.50,'已发货','2024-08-23 14:40:45'),(104,16,120.20,'待支付','2024-08-23 14:43:00'),(105,12,135.50,'已支付','2024-08-23 14:45:15'),(106,3,145.20,'待支付','2024-08-23 14:47:30'),(107,8,130.20,'已发货','2024-08-23 14:49:45'),(108,19,125.50,'已支付','2024-08-23 14:52:00'),(109,6,135.20,'待支付','2024-08-23 14:54:15'),(110,14,140.50,'已发货','2024-08-23 14:56:30'),(111,10,120.20,'待支付','2024-08-23 14:58:45'),(112,13,135.50,'已支付','2024-08-23 15:01:00'),(113,18,145.20,'待支付','2024-08-23 15:03:15'),(114,6,130.20,'已发货','2024-08-23 15:05:30'),(115,11,125.50,'已支付','2024-08-23 15:07:45'),(116,19,135.20,'待支付','2024-08-23 15:10:00'),(117,5,140.50,'已发货','2024-08-23 15:12:15'),(118,20,120.20,'待支付','2024-08-23 15:14:30'),(119,17,135.50,'已支付','2024-08-23 15:16:45'),(120,4,145.20,'待支付','2024-08-23 15:19:00'),(121,10,130.20,'已发货','2024-08-23 15:21:15'),(122,13,125.50,'已支付','2024-08-23 15:23:30'),(123,18,135.20,'待支付','2024-08-23 15:25:45'),(124,6,140.50,'已发货','2024-08-23 15:28:00'),(125,11,120.20,'待支付','2024-08-23 15:30:15'),(126,19,135.50,'已支付','2024-08-23 15:32:30'),(127,5,145.20,'待支付','2024-08-23 15:34:45'),(128,20,130.20,'已发货','2024-08-23 15:37:00'),(129,17,125.50,'已支付','2024-08-23 15:39:15'),(130,4,135.20,'待支付','2024-08-23 15:41:30'),(131,10,140.50,'已发货','2024-08-23 15:43:45'),(132,13,120.20,'待支付','2024-08-23 15:46:00'),(133,18,135.50,'已支付','2024-08-23 15:48:15'),(134,6,145.20,'待支付','2024-08-23 15:50:30'),(135,11,130.20,'已发货','2024-08-23 15:52:45'),(136,19,125.50,'已支付','2024-08-23 15:55:00'),(137,5,135.20,'待支付','2024-08-23 15:57:15'),(138,20,140.50,'已发货','2024-08-23 15:59:30'),(139,17,120.20,'待支付','2024-08-23 16:01:45'),(140,4,135.50,'已支付','2024-08-23 16:04:00'),(141,10,145.20,'待支付','2024-08-23 16:06:15'),(142,13,130.20,'已发货','2024-08-23 16:08:30'),(143,18,125.50,'已支付','2024-08-23 16:10:45'),(144,6,135.20,'待支付','2024-08-23 16:13:00'),(145,11,140.50,'已发货','2024-08-23 16:15:15'),(146,19,120.20,'待支付','2024-08-23 16:17:30'),(147,5,135.50,'已支付','2024-08-23 16:19:45'),(148,20,145.20,'待支付','2024-08-23 16:22:00'),(149,17,130.20,'已发货','2024-08-23 16:24:15'),(150,4,125.50,'已支付','2024-08-23 16:26:30'),(151,10,135.20,'待支付','2024-08-23 16:28:45'),(152,13,140.50,'已发货','2024-08-23 16:31:00'),(153,18,120.20,'待支付','2024-08-23 16:33:15'),(154,6,135.50,'已支付','2024-08-23 16:35:30'),(155,11,145.20,'待支付','2024-08-23 16:37:45'),(156,19,130.20,'已发货','2024-08-23 16:40:00'),(157,5,125.50,'已支付','2024-08-23 16:42:15'),(158,20,135.20,'待支付','2024-08-23 16:44:30'),(159,17,140.50,'已发货','2024-08-23 16:46:45'),(160,4,120.20,'待支付','2024-08-23 16:49:00'),(161,10,135.50,'已支付','2024-08-23 16:51:15'),(162,13,145.20,'待支付','2024-08-23 16:53:30'),(163,18,130.20,'已发货','2024-08-23 16:55:45'),(164,6,125.50,'已支付','2024-08-23 16:58:00'),(165,11,135.20,'待支付','2024-08-23 17:00:15'),(166,19,140.50,'已发货','2024-08-23 17:02:30'),(167,5,120.20,'待支付','2024-08-23 17:04:45'),(168,20,135.50,'已支付','2024-08-23 17:07:00'),(169,17,145.20,'待支付','2024-08-23 17:09:15'),(170,4,130.20,'已发货','2024-08-23 17:11:30'),(171,10,125.50,'已支付','2024-08-23 17:13:45'),(172,13,135.20,'待支付','2024-08-23 17:16:00'),(173,18,140.50,'已发货','2024-08-23 17:18:15'),(174,6,120.20,'待支付','2024-08-23 17:20:30'),(175,11,135.50,'已支付','2024-08-23 17:22:45'),(176,19,145.20,'待支付','2024-08-23 17:25:00'),(177,5,130.20,'已发货','2024-08-23 17:27:15'),(178,20,125.50,'已支付','2024-08-23 17:29:30'),(179,17,135.20,'待支付','2024-08-23 17:31:45'),(180,4,140.50,'已发货','2024-08-23 17:34:00'),(181,10,120.20,'待支付','2024-08-23 17:36:15'),(182,13,135.50,'已支付','2024-08-23 17:38:30'),(183,18,145.20,'待支付','2024-08-23 17:40:45'),(184,6,130.20,'已发货','2024-08-23 17:43:00'),(185,11,125.50,'已支付','2024-08-23 17:45:15'),(186,19,135.20,'待支付','2024-08-23 17:47:30'),(187,5,140.50,'已发货','2024-08-23 17:49:45'),(188,20,120.20,'待支付','2024-08-23 17:52:00'),(189,17,135.50,'已支付','2024-08-23 17:54:15'),(190,4,145.20,'待支付','2024-08-23 17:56:30'),(191,10,130.20,'已发货','2024-08-23 17:58:45'),(192,13,125.50,'已支付','2024-08-23 18:01:00'),(193,18,135.20,'待支付','2024-08-23 18:03:15'),(194,6,140.50,'已发货','2024-08-23 18:05:30'),(195,11,120.20,'待支付','2024-08-23 18:07:45'),(196,19,135.50,'已支付','2024-08-23 18:10:00'),(197,5,145.20,'待支付','2024-08-23 18:12:15'),(198,20,130.20,'已发货','2024-08-23 18:14:30'),(199,17,125.50,'已支付','2024-08-23 18:16:45'),(200,4,135.20,'待支付','2024-08-23 18:19:00'),(201,10,140.50,'已发货','2024-08-23 18:21:15'),(202,13,120.20,'待支付','2024-08-23 18:23:30'),(203,18,135.50,'已支付','2024-08-23 18:25:45'),(204,6,145.20,'待支付','2024-08-23 18:28:00'),(205,11,130.20,'已发货','2024-08-23 18:30:15'),(206,19,125.50,'已支付','2024-08-23 18:32:30'),(207,5,135.20,'待支付','2024-08-23 18:34:45'),(208,20,140.50,'已发货','2024-08-23 18:37:00'),(209,17,120.20,'待支付','2024-08-23 18:39:15'),(210,4,135.50,'已支付','2024-08-23 18:41:30'),(211,10,145.20,'待支付','2024-08-23 18:43:45');
部署HAI高算力服务器（大语言模型）
python环境配置
需要的python库
pip install openai
pip install langchain
pip install langchain-core
pip install langchain-community
pip install mysql-connector-python
pip install streamlit
pip install plotly
pip install numpy
pip install pandas
pip install watchdog
pip install matplotlib
pip install kaleido
搭建项目框架
新建名为 workspace 文件夹进行保存项目代码
在项目文件夹（workspace）中新建配置文件 config.yaml
在项目文件夹（workspace）中新建应用主文件 text2sql2plotly.py
打开 config.yaml 文件，复制以下内容到配置文件中在这里插入代码片
这里主要分为 database 配置 和 hai 的配置
database 的配置详解：
db_user: 数据库账号，默认为 root
db_password: 创建数据库时的密码
db_host: 数据库连接地址
db_port: 数据库公网端口
db_name 创建的数据库名称，如果按手册来默认是 shop
hai 配置详解：
model 使用的大模型
base_url 模型暴露的 api 地址，是公网 ip 和端口的组合，默认 llama端口是6399
database 中填入 TDSQL-C 的相关配置，db_host、db_port可以在集群列表中找到
<code>database:
db_user: root
db_password: tencent_TDSQL
db_host: sh-cynosdbmysql-grp-9d8prc9o.sql.tencentcdb.com
db_port: 21919
db_name: shop
hai:
model: llama3.1:8b
base_url: http://82.156.229.112:6399
将以下程序代码复制并保存到 text2sql2plotly.py 文件中
from langchain_community.utilities import SQLDatabase
from langchain_core.prompts import ChatPromptTemplate
from langchain_community.chat_models import ChatOllama
from langchain_core.output_parsers import StrOutputParser
from langchain_core.runnables import RunnablePassthrough
import yaml
import mysql.connector
from decimal import Decimal
import plotly.graph_objects as go
import plotly
import pkg_resources
import matplotlib
yaml_file_path = 'config.yaml'
with open(yaml_file_path, 'r') as file:
config_data = yaml.safe_load(file)
#获取所有的已安装的pip包
def get_piplist(p):
return [d.project_name for d in pkg_resources.working_set]
#获取llm用于提供AI交互
ollama = ChatOllama(model=config_data['hai']['model'],base_url=config_data['hai']['base_url'])
db_user = config_data['database']['db_user']
db_password = config_data['database']['db_password']
db_host = config_data['database']['db_host']
db_port= config_data['database']['db_port']
db_name = config_data['database']['db_name']
# 获得schema
def get_schema(db):
schema = mysql_db.get_table_info()
return schema
def getResult(content):
global mysql_db
# 数据库连接
mysql_db = SQLDatabase.from_uri(f"mysql+mysqlconnector://{ db_user}:{ db_password}@{ db_host}:{ db_port}/{ db_name}")
# 获得 数据库中表的信息
#mysql_db_schema = mysql_db.get_table_info()
#print(mysql_db_schema)
template = """基于下面提供的数据库schema, 根据用户提供的要求编写sql查询语句,要求尽量使用最优sql,每次查询都是独立的问题，不要收到其他查询的干扰:
{schema}
Question: {question}
只返回sql语句，不要任何其他多余的字符，例如markdown的格式字符等:
如果有异常抛出不要显示出来
"""
prompt = ChatPromptTemplate.from_template(template)
text_2_sql_chain = (
RunnablePassthrough.assign(schema=get_schema)
| prompt
| ollama
| StrOutputParser()
)
# 执行langchain 获取操作的sql语句
sql = text_2_sql_chain.invoke({ "question": content})
print(sql)
#连接数据库进行数据的获取
# 配置连接信息
conn = mysql.connector.connect(
host=db_host,
port=db_port,
user=db_user,
password=db_password,
database=db_name
)
# 创建游标对象
cursor = conn.cursor()
# 查询数据
cursor.execute(sql.strip("```").strip("```sql"))
info = cursor.fetchall()
# 打印结果
#for row in info:
#print(row)
# 关闭游标和数据库连接
cursor.close()
conn.close()
#根据数据生成对应的图表
print(info)
template2 = """
以下提供当前python环境已经安装的pip包集合:
{installed_packages};
请根据data提供的信息，生成是一个适合展示数据的plotly的图表的可执行代码,要求如下：
1.不要导入没有安装的pip包代码
2.如果存在多个数据类别，尽量使用柱状图，循环生成时图表中对不同数据请使用不同颜色区分，
3.图表要生成图片格式,保存在当前文件夹下即可，名称固定为：图表.png,
4.我需要您生成的代码是没有 Markdown 标记的，纯粹的编程语言代码。
5.生成的代码请注意将所有依赖包提前导入，
6.不要使用iplot等需要特定环境的代码
7.请注意数据之间是否可以转换，使用正确的代码
8.不需要生成注释
data:{data}
这是查询的sql语句与文本:
sql:{sql}
question:{question}
返回数据要求:
仅仅返回python代码，不要有额外的字符
"""
prompt2 = ChatPromptTemplate.from_template(template2)
data_2_code_chain = (
RunnablePassthrough.assign(installed_packages=get_piplist)
| prompt2
| ollama
| StrOutputParser()
)
# 执行langchain 获取操作的sql语句
code = data_2_code_chain.invoke({ "data": info,"sql":sql,'question':content})
#删除数据两端可能存在的markdown格式
print(code.strip("```").strip("```python"))
exec(code.strip("```").strip("```python"))
return { "code":code,"SQL":sql,"Query":info}
# 构建展示页面
import streamlit
# 设置页面标题
streamlit.title('AI驱动的数据库TDSQL-C 电商可视化分析小助手')
# 设置对话框
content = streamlit.text_area('请输入想查询的信息', value='', max_chars=None)code>
# 提问按钮 # 设置点击操作
if streamlit.button('提问'):
#开始ai及langchain操作
if content:
#进行结果获取
result = getResult(content)
#显示操作结果
streamlit.write('AI生成的SQL语句:')
streamlit.write(result['SQL'])
streamlit.write('SQL语句的查询结果:')
streamlit.write(result['Query'])
streamlit.write('plotly图表代码:')
streamlit.write(result['code'])
# 显示图表内容（生成在getResult中）
streamlit.image('./图表.png', width=800)
打开终端执行以下命令
streamlit run text2sql2plotly.py
命令运行后在浏览器中打开UI界面
例如：
API支持模型类型
输入：查询一下每类商品的名称和对应的销售总额 测试效果
声明
本文内容仅代表作者观点，或转载于其他网站，本站不以此文作为商业用途
如有涉及侵权，请联系本站进行删除
转载本站原创文章，请注明来源及作者。