Snowflake Magic

An ipython magic function to simplify usage of Snowflake SQL in your notebooks.

Example

import pandas as pd
result = %snowflake SELECT timestamp, value FROM mytable;
df = pd.DataFrame(result)
df.plot.line();

Setup and Configuration

Install the extension

pip install snowflakemagic

Load extension

%reload_ext snowflakemagic

Available magic functions

%snowflake_auth

Inline function connecting to your snowflake account. Reads connection parameters from .env file:

You can either authenticate via SSO, which opens an external browser, or using credentials.

Provide your snowflake account details:

snowflake_account="<YOUR-SNOWFLAKE-ACCOUNT>"

If you want to connect via sso, provide your sso username:

snowflake_ssouser="<YOUR-SSO-USERNAME>"

If you want to connect via use-credentials, provide the password, otherwise SSO authentication is used.

snowflake_user="<YOUR-USERNAME>"
snowflake_password="<YOUR-PASSWORD>"

For more details on .env file see How to NOT embedded credential in Jupyter notebook or python-dotenv

%%snowflake, %snowflake or %snowflake_script

• Executes a snowflake query/script and returns the result as a json object.
• Multiple queries/statements separated by ; can be exceuted, but only last result will be returned.
• A query MUST end with a semi-colon (;)

Example 1

Query in code-cell

%%snowflake my_result
SELECT * 
    FROM xyz;

.. use result in another code cell:

import pandas as pd

#put result into a dataframe
df = pd.DataFrame(my_result)

#...

Example 2 - Inline query

import pandas as pd

my_result = %snowflake SELECT * FROM xyz;
df = pd.DataFrame(my_result)

#...

Example 3 - From script

Query using external query script files e.g. myscript.snowql

SELECT * FROM xyz;

Then in your code-cell, pass the script name

import pandas as pd

my_result = %snowflake_script myscript.snowql
df = pd.DataFrame(my_result)

#...

Example 4 - Parameterized script

Query using external query script files e.g. myscript.snowql which can be parameterized

SELECT * FROM xyz WHERE mycolumn=@MYVALUE@@;

Then in your code-cell, pass the script name

import pandas as pd

my_result = %snowflake_script myscript.snowql @@MYVALUE@@=test
df = pd.DataFrame(my_result)

#...

Example 5 - Chaining multiple scripts

You can also chain multiple scripts

mycte.snowql

WITH
    my_cte AS (
        SELECT col_1, col_2
            FROM xyz
    )

myscript.snowql

SELECT * FROM my_cte WHERE col_1=@MYVALUE@@;

Then in your code-cell, you can append the various script files

import pandas as pd

my_result = %snowflake_script mycte.snowql<<myscript.snowql @@MYVALUE@@=test
df = pd.DataFrame(my_result)

#...

Getting Started

If you are new to Jupyter notebooks, I recommend getting started using Jupyter notebooks in Visual Studio Code

1. Configure .env file providing connection parameters as explained above
2. Start using getting-started.ipynp to learn how to use the magic functions