Samples on Azure Data Science Virtual Machines

Azure Data Science Virtual Machines (DSVMs) include a comprehensive set of sample code. These samples include Jupyter notebooks and scripts in languages like Python and R.


For more information about how to run Jupyter notebooks on your data science virtual machines, see the Access Jupyter section.


In order to run these samples, you must have provisioned a Data Science Virtual Machine. See the quickstarts for Windows and Ubuntu.

Available samples

Samples category Description Locations
R language Samples illustrate scenarios such as how to connect with Azure-based cloud data stores and how to compare open-source R and Microsoft Machine Learning Server. They also explain how to operationalize models on Microsoft Machine Learning Server and SQL Server.
R language





Python language Samples explain scenarios like how to connect with Azure-based cloud data stores and how to work with Azure Machine Learning.
Python language


Julia language Provides a detailed description of plotting and deep learning in Julia. Also explains how to call C and Python from Julia.
Julia language



Azure Machine Learning Illustrates how to build machine-learning and deep-learning models with Machine Learning. Deploy models anywhere. Use automated machine learning and intelligent hyperparameter tuning. Also use model management and distributed training.
Machine Learning


PyTorch notebooks Deep-learning samples that use PyTorch-based neural networks. Notebooks range from beginner to advanced scenarios.
PyTorch notebooks


TensorFlow A variety of neural network samples and techniques implemented by using the TensorFlow framework.


Microsoft Cognitive Toolkit
Deep-learning samples published by the Cognitive Toolkit team at Microsoft.
Cognitive Toolkit



Caffe2 Deep-learning samples that use Caffe2-based neural networks. Several notebooks familiarize users with Caffe2 and how to use it effectively. Examples include image preprocessing and dataset creation. They also include regression and how to use pretrained models.


H2O Python-based samples that use H2O for real-world problem scenarios.


SparkML language Samples that use features of the Apache Spark MLLib toolkit through pySpark and MMLSpark: Microsoft Machine Learning for Apache Spark on Apache Spark 2.x.
SparkML language


XGBoost Standard machine-learning samples in XGBoost for scenarios like classification and regression.


Access Jupyter

To access Jupyter, select the Jupyter icon on the desktop or application menu. You also can access Jupyter on a Linux edition of a DSVM. To access remotely from a web browser, go to https://<Full Domain Name or IP Address of the DSVM>:8000 on Ubuntu.

To add exceptions and make Jupyter access available over a browser, use the following guidance:

Enable Jupyter exception

Sign in with the same password that you use to log in to the Data Science Virtual Machine.

Jupyter home
Jupyter home

R language

R samples

Python language

Python samples

Julia language

Julia samples

Azure Machine Learning

Azure Machine Learning samples


PyTorch samples


TensorFlow samples


CNTK samples


caffe2 samples


H2O samples


SparkML samples


XGBoost samples