Python和Snowflake使用SQL Alchemy引擎将新数据追加到现有表中，返回的是当前会话，但没有当前数据库

我需要在snowflake上的现有表中添加一些新数据。我正在使用

sqlalchemy

作为引擎，同时使用pandas数据帧

来创建sql（）

。以下是导入和脚本：

import pandas as pd
import os
import snowflake.connector as snowCtx
import getpass
import json
import numpy as np
from datetime import date, datetime
import time
from sqlalchemy import create_engine
from sqlalchemy.dialects import registry
import snowflake.sqlalchemy
from snowflake.connector.pandas_tools import pd_writer
from sqlalchemy.ext.declarative import declarative_base

registry.register('snowflake', 'snowflake.sqlalchemy', 'dialect')

columns_df = pd.DataFrame(data.columns.to_list(), columns={'survey_column_name'})
                        columns_df['survey_id'] = nextval
                        columns_df = columns_df[['survey_id', 'survey_column_name']]
                        columns_df.to_sql('SURVEY_METADATA_COLUMN_NAMES', 
                                         index = False,  
                                         index_label = None, 
                                         con = engine, 
                                         schema = 'PUBLIC', 
                                         if_exists = 'append', 
                                         chunksize = 300,
                                         method = pd_writer)

我得到的错误如下：

user = input('Your Snowflake username: ')
password = getpass.getpass('Your Snowflake Password: ')
account = 'MY_ACCOUNT'
conn = snowCtx.connect(
    user=user,
    password=password,
    account=account,
    database='MY_DB',
    schema='PUBLIC',
    warehouse='COMPUTE_WH',
    role='SYSADMIN'
)

engine = create_engine(
    'snowflake://{user}:{password}@{account}/'.format(
        user=user,
        password=password,
        account=account,
        database='MY_DB',
        schema = 'PUBLIC',
        warehouse='COMPUTE_WH',
        role='SYSADMIN',
        cache_column_metadata=True
    )
)

编程错误：（雪花。连接器。错误。编程错误）090105 （22000）：无法执行创建

桌子。此会话没有当前数据库。调用“使用” “数据库”，或使用限定名称。[SQL:

创建表“PUBLIC”。“调查\元数据\列\名称”(

调查id整数

调查（列名称文本）

]

连接如下：

user = input('Your Snowflake username: ')
password = getpass.getpass('Your Snowflake Password: ')
account = 'MY_ACCOUNT'
conn = snowCtx.connect(
    user=user,
    password=password,
    account=account,
    database='MY_DB',
    schema='PUBLIC',
    warehouse='COMPUTE_WH',
    role='SYSADMIN'
)

engine = create_engine(
    'snowflake://{user}:{password}@{account}/'.format(
        user=user,
        password=password,
        account=account,
        database='MY_DB',
        schema = 'PUBLIC',
        warehouse='COMPUTE_WH',
        role='SYSADMIN',
        cache_column_metadata=True
    )
)

---

我改为使用

write\u pandas（）

：

success, nchunks, nrows, _ = write_pandas(conn, 
                                          columns_df, 
                                          'SURVEY_METADATA_COLUMN_NAMES', 
                                          chunk_size = 300, 
                                          schema = 'PUBLIC')
                        print(success, nchunks, nrows)
if(success):
   print(filename+' columns uploaded')
else:
   print(filename+' columns were not uploaded')

它需要

pyarrow

库，因此我使用以下方法安装它：

pip install pyarrow

我删除了与

sqlalchemy

相关的所有导入，并保留了以下内容：

import pandas as pd
import os
import snowflake.connector as snowCtx
import getpass
import json
import numpy as np
from datetime import date, datetime
import time
from snowflake.connector.pandas_tools import write_pandas