Python Tutorial - Deploy Model - SQL Machine Learning - Microsoft Docs
Python Tutorial - Deploy Model - SQL Machine Learning - Microsoft Docs
Python Tutorial - Deploy Model - SQL Machine Learning - Microsoft Docs
In this article
Prerequisites
Create a stored procedure that generates the model
Store the model in a database table
Create a stored procedure that makes predictions
Next steps
Applies to: SQL Server 2017 (14.x) and later Azure SQL Managed Instance
In part four of this four-part tutorial series, you'll deploy a linear regression model developed in Python into a SQL Server database
using Machine Learning Services or Big Data Clusters.
In part two, you learned how to load the data from a database into a Python data frame, and prepare the data in Python.
In part three, you learned how to train a linear regression machine learning model in Python.
https://docs.microsoft.com/en-us/sql/machine-learning/tutorials/python-ski-rental-linear-regression-deploy-model?view=sql-server-ver15 1/8
4/5/2021 Python tutorial: Deploy model - SQL machine learning | Microsoft Docs
Prerequisites
Part four of this tutorial assumes you have completed part one and its prerequisites.
Run the following T-SQL statement in Azure Data Studio to create the stored procedure to train the model.
SQL = Copy
-- Stored procedure that trains and generates a Python model using the rental_data and a linear regression algorithm
DROP PROCEDURE IF EXISTS generate_rental_py_model;
go
CREATE PROCEDURE generate_rental_py_model (@trained_model varbinary(max) OUTPUT)
AS
BEGIN
EXECUTE sp_execute_external_script
@language = N'Python'
, @script = N'
from sklearn.linear_model import LinearRegression
import pickle
df = rental_train_data
https://docs.microsoft.com/en-us/sql/machine-learning/tutorials/python-ski-rental-linear-regression-deploy-model?view=sql-server-ver15 2/8
4/5/2021 Python tutorial: Deploy model - SQL machine learning | Microsoft Docs
lin_model = LinearRegression()
, @input_data_1 = N'select "RentalCount", "Year", "Month", "Day", "WeekDay", "Snow", "Holiday" from dbo.rental_data
where Year < 2015'
, @input_data_1_name = N'rental_train_data'
, @params = N'@trained_model varbinary(max) OUTPUT'
, @trained_model = @trained_model OUTPUT;
END;
GO
1. Run the following T-SQL statement in Azure Data Studio to create a table called dbo.rental_py_models which is used to store the
model.
SQL = Copy
USE TutorialDB;
DROP TABLE IF EXISTS dbo.rental_py_models;
GO
CREATE TABLE dbo.rental_py_models (
model_name VARCHAR(30) NOT NULL DEFAULT('default model') PRIMARY KEY,
model VARBINARY(MAX) NOT NULL
);
GO
https://docs.microsoft.com/en-us/sql/machine-learning/tutorials/python-ski-rental-linear-regression-deploy-model?view=sql-server-ver15 3/8
4/5/2021 Python tutorial: Deploy model - SQL machine learning | Microsoft Docs
2. Save the model to the table as a binary object, with the model name linear_model.
SQL = Copy
SQL = Copy
EXECUTE sp_execute_external_script
@language = N'Python',
@script = N'
rental_model = pickle.loads(py_model)
https://docs.microsoft.com/en-us/sql/machine-learning/tutorials/python-ski-rental-linear-regression-deploy-model?view=sql-server-ver15 4/8
4/5/2021 Python tutorial: Deploy model - SQL machine learning | Microsoft Docs
df = rental_score_data
# Compute error between the test predictions and the actual values.
lin_mse = mean_squared_error(lin_predictions, df[target])
#print(lin_mse)
predictions_df = pandas.DataFrame(lin_predictions)
END;
GO
SQL = Copy
SQL = Copy
--Insert the results of the predictions for test set into a table
INSERT INTO py_rental_predictions
EXEC py_predict_rentalcount 'linear_model';
https://docs.microsoft.com/en-us/sql/machine-learning/tutorials/python-ski-rental-linear-regression-deploy-model?view=sql-server-ver15 6/8
4/5/2021 Python tutorial: Deploy model - SQL machine learning | Microsoft Docs
You have successfully created, trained, and deployed a model. You then used that model in a stored procedure to predict values based
on new data.
Next steps
In part four of this tutorial series, you completed these steps:
To learn more about using Python with SQL machine learning, see:
Python tutorials
https://docs.microsoft.com/en-us/sql/machine-learning/tutorials/python-ski-rental-linear-regression-deploy-model?view=sql-server-ver15 7/8
4/5/2021 Python tutorial: Deploy model - SQL machine learning | Microsoft Docs
Recommended content
Python tutorial: Train model - SQL Python tutorial: Predict NYC taxi Python tutorial: Prepare data - SQL
machine learning fares with binary classification - machine learning
In part three of this four-part tutorial SQL machine learning In part two of this four-part tutorial
series, you'll train a linear regression Learn how to embed Python code in SQL series, you'll use Python to prepare data
model in Python to predict ski rentals… Server stored procedures and T-SQL to predict ski rentals with SQL machine…
functions with SQL machine learning to…
Show more S
https://docs.microsoft.com/en-us/sql/machine-learning/tutorials/python-ski-rental-linear-regression-deploy-model?view=sql-server-ver15 8/8