"No driver name specified" writing pandas data frame into SQL Server table

Question

I am trying to write a Pandas' DataFrame into an SQL Server table. Here is my example:

import pyodbc
import pandas as pd
import sqlalchemy

df = pd.DataFrame({'MDN': [242342342] })
engine = sqlalchemy.create_engine('mssql://localhost/Sandbox?trusted_connection=yes')
df.to_sql('Test',engine, if_exists = 'append',index = False)

I am getting the following error message. Any thoughts on how to fix?

c:\python34\lib\site-packages\sqlalchemy\connectors\pyodbc.py:82: SAWarning: No driver name specified; this is expected by PyODBC when using DSN-less connections
  "No driver name specified; "


---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-25-78677a18ce2d> in <module>()
      4 engine = sqlalchemy.create_engine('mssql://localhost/Sandbox?trusted_connection=yes')
      5 
----> 6 df.to_sql('Test',engine, if_exists = 'append',index = False)
      7 
      8 #cnxn.close()

c:\python34\lib\site-packages\pandas\core\generic.py in to_sql(self, name, con, flavor, schema, if_exists, index, index_label, chunksize, dtype)
    980             self, name, con, flavor=flavor, schema=schema, if_exists=if_exists,
    981             index=index, index_label=index_label, chunksize=chunksize,
--> 982             dtype=dtype)
    983 
    984     def to_pickle(self, path):

c:\python34\lib\site-packages\pandas\io\sql.py in to_sql(frame, name, con, flavor, schema, if_exists, index, index_label, chunksize, dtype)
    547     pandas_sql.to_sql(frame, name, if_exists=if_exists, index=index,
    548                       index_label=index_label, schema=schema,
--> 549                       chunksize=chunksize, dtype=dtype)
    550 
    551 

c:\python34\lib\site-packages\pandas\io\sql.py in to_sql(self, frame, name, if_exists, index, index_label, schema, chunksize, dtype)
   1564                             if_exists=if_exists, index_label=index_label,
   1565                             dtype=dtype)
-> 1566         table.create()
   1567         table.insert(chunksize)
   1568 

c:\python34\lib\site-packages\pandas\io\sql.py in create(self)
    646 
    647     def create(self):
--> 648         if self.exists():
    649             if self.if_exists == 'fail':
    650                 raise ValueError("Table '%s' already exists." % self.name)

c:\python34\lib\site-packages\pandas\io\sql.py in exists(self)
    634 
    635     def exists(self):
--> 636         return self.pd_sql.has_table(self.name, self.schema)
    637 
    638     def sql_schema(self):

c:\python34\lib\site-packages\pandas\io\sql.py in has_table(self, name, schema)
   1577         query = flavor_map.get(self.flavor)
   1578 
-> 1579         return len(self.execute(query, [name,]).fetchall()) > 0
   1580 
   1581     def get_table(self, table_name, schema=None):

c:\python34\lib\site-packages\pandas\io\sql.py in execute(self, *args, **kwargs)
   1465             cur = self.con
   1466         else:
-> 1467             cur = self.con.cursor()
   1468         try:
   1469             if kwargs:

AttributeError: 'Engine' object has no attribute 'cursor'

Also, is there ways to write connection string for create_engine differently? I would love to write it in form of a dictionary rather than a string.

Update : Here is my new environment:

MS SQL Server: Microsoft SQL Server 2012 - 11.0.2100.60 (X64) Feb 10 2012 19:39:15 Copyright (c) Microsoft Corporation Standard Edition (64-bit) on Windows NT 6.2 (Build 9200: )

Python: 3.4.3 (v3.4.3:9b73f1c3e601, Feb 24 2015, 22:43:06) [MSC v.1600 32 bit (Intel)]

Pandas version: '0.16.2'

sqlalchemy version: 1.1.3

Jupyter server version: 4.2.3

Now the line

engine = sqlalchemy.create_engine('mssql+pyodbc://localhost/Sandbox?trusted_connection=yes')

generates the following error:

c:\python34\lib\site-packages\sqlalchemy\connectors\pyodbc.py:82: SAWarning: No driver name specified; this is expected by PyODBC when using DSN-less connections
  "No driver name specified; "

Answer 1

You need to specify both that you want to use ODBC and what ODBC driver to use.

engine = sqlalchemy.create_engine('mssql+pyodbc://localhost/Sandbox?driver=SQL+Server+Native+Client+11.0')

Trusted connections are the default, so you don't need to specify that, although it shouldn't hurt to do so.

Answer 2

The likely problem is that you have not specified the driver, so try:

engine = sqlalchemy.create_engine('mssql+pyodbc://localhost/Sandbox?trusted_connection=yes')

This is based on the warning message that you got on the top:

c:\python34\lib\site-packages\sqlalchemy\connectors\pyodbc.py:82: SAWarning: No driver name specified; this is expected by PyODBC when using DSN-less connections
  "No driver name specified; "

Note that you can also use pymssql instead of pyodbc, but MS recommends the latter.

---

EDIT

---

Here is official documentation on how to connect with/without DSN (data source name):

https://github.com/mkleehammer/pyodbc/blob/master/docs/index.md#connect-to-a-database

Answer 3

I know the question has been answered for some time now and it's just a warning, but if you have transferred everything correctly and this error still occurs it's annoying.

For all those who had to struggle with it like I did, you can also enter the driver directly in the script, Pyodbc.py offers the possibility for this (row 26 - 28):

    # for non-DSN connections, this *may* be used to
    # hold the desired driver name
    pyodbc_driver_name = 'ODBC Driver 17 for SQL Server'

Answer 4

Above information was much useful. Commenting below version of mine as consolidated which can help freshers during search.

#using library pandas and pyodbc - if not available please use pip install commands to install library based on version. Python version used here is 3.7.8

import pandas as pd
from sqlalchemy import create_engine
import pyodbc
 
  
#This query will work for sql authentication
def mssql_engine():
    engine = create_engine('mssql+pyodbc://type_username:type_password@type_servername_or_localhostname/type_database_name?driver=SQL+Server+Native+Client+11.0')
    return engine

#This query will for windows authentication 
#Note: Uncomment below code for windows authentication
#def mssql_engine():
      #engine = create_engine('mssql+pyodbc://localhostname/db_name?driver=SQL+Server+Native+Client+11.0')
      #return engine
   
 
query = 'select * from table_name'
#using pandas to read from sql and passing connection string as function
df = pd.read_sql(query, mssql_engine() ) 

#printing result
print(df)

Answer 5

I ran into a similar problem with SQLModel with an error message like:

SAWarning: No driver name specified; this is expected by PyODBC when using DSN-less connections return _create_engine(url, **current_kwargs) # type: ignore

Using @cco 's response solved the issue:

CONNECTION_URL = f"mssql+pyodbc://{SQL_SERVER_USERNAME}:{SQL_SERVER_PASSWORD}" \
                    f"@{SQL_SERVER_NAME}/{SQL_SERVER_DATABASE}?driver=SQL%20Server"