设置 Dataframe loc 导致 SQL 字符串 - 元组错误

Question

当我在 pd.read_sql_query 字符串中使用 df.loc 值时出现元组错误。

import pyodbc as db
import pandas as pd

#Connection String
Conn = db.connect('Driver={driver};Server=server;Database=db;Trusted_Connection=yes')

#DataFrame that stores SQL LookUp Table
dfSQLLookUp = pd.read_sql_query('Select col1, 
col2 from dbo.LookUp_TableNames', Conn)

#Loop to store table in Dataframe and (eventually) ouput to Excel file
varIndex = -1
for i in dfSQLLookUp.itertuples():
varIndex += 1
dfResults = pd.read_sql_query('if (exists (select * from information_schema.TABLES where TABLE_SCHEMA = dbo and TABLE_NAME = {})) select * from dbo.{} else print table does not exist'.format(dfSQLLookUp.loc[varIndex, 'SQLTableNameColumn']), Conn)

当我打印 df.loc 值时：

import pyodbc as db
import pandas as pd

#Connection String
Conn = db.connect('Driver={driver};Server=server;Database=db;Trusted_Connection=yes')

#DataFrame that stores SQL LookUp Table
dfSQLLookUp = pd.read_sql_query('Select col1, 
col2 from dbo.LookUp_TableNames', Conn)

varIndex = -1
for i in dfSQLNames.itertuples():
    varIndex += 1
    print (dfSQLNames.loc[varIndex, 'IFSFlatFileName'])

返回了正确的字符串值。这是 pd.read_sql_query 语法的问题吗？还是元组错误？感谢任何见解。

元组错误文本：

dfResults = pd.read_sql_query('if (exists (select * from information_schema.TABLES where TABLE_SCHEMA = dbo and TABLE_NAME = {})) select * from dbo.{} else print table does not exist'.format(dfSQLLookUp.loc[varIndex, 'SQLTableNamesColumn']), Conn) IndexError: tuple index out of range Press any key to continue . . .

Answer 1

正如@MaxU 在评论中提到的，Pandas / SQL Alchemy 不支持 T-SQL。这很可能是引发元组错误的原因。为了绕过 Pandas / SQL Alchemy 的限制，我在 SQL 中创建了一个存储过程来填充查找 table。我还必须在过程中设置 NOCOUNT ON，因为在 Python 中调用过程时出现 NullType 错误。下面是最终结果。希望这可以帮助其他人代替我。

import pyodbc as db
import pandas as pd

#Database Connection String
Conn = db.connect('Driver={driver};Server=server;Database=db;Trusted_Connection=yes')

#DataFrame that stores the LookUp values for the SQL table names and the Excel file names
dfSQLLookUp = pd.read_sql_query('Exec dbo.uspDataFrameLookUp', Conn)

#Loop to store SQL tables in a Dataframe and output to Excel file
varIndex = -1
for i in dfSQLLookUp.itertuples():
    varIndex += 1
    dfResults = pd.read_sql_query('select * from dbo.{}'.format(dfSQLLookUp.loc[varIndex, 'NameOfCol1']), Conn)

    #Writing the Dataframe to an Excel file
    OutputDirectory = r'\path\...\PythonExtracts\{}.xlsx'.format(dfSQLLookUp.loc[varIndex, 'NameOfCol2'])
    Excel = pd.ExcelWriter(OutputDirectory, engine='xlsxwriter')
    dfResults.to_excel(Excel, sheet_name='Sheet1')
    Excel.save()

Conn.close()

存储过程

ALTER procedure [dbo].[uspDataFrameLookUp]
as

begin

set nocount on

declare @varTableName nvarchar(30)
declare @varCountRows int
declare @varCountLoop int
declare @varDFTable table (NameOfCol1 nvarchar(50), 
NameOfCol2 nvarchar(50))

set @varCountRows = (select count(*) from [dbo].[LookUp_TableNames])
set @varCountLoop = 0

while @varCountLoop <= @varCountRows
begin
set @varCountLoop = @varCountLoop + 1
set @varTableName = (select NameOfCol1 from 
dbo.LookUp_TableNames where ID = @varCountLoop)
if (exists (select * 
                from information_schema.TABLES 
                where TABLE_SCHEMA = 'dbo' 
                and TABLE_NAME = @varTableName))

    insert into @varDFTable 
            select NameOfCol1, NameOfCol2
            from dbo.LookUp_TableNames 
            where NameOfCol1 = @varTableName
end

select * from @varDFTable order by NameOfCol1 asc

end

设置 Dataframe loc 导致 SQL 字符串 - 元组错误

Setting Dataframe loc result in SQL string - Tuples Error

python

pyodbc

dataframe

python-3.x

pandas