azureml-dataprep 1.1.10

Azure ML Data Preparation SDK

Azure Machine Learning Data Prep SDK for Python

The Azure Machine Learning Data Prep SDK is used to load, transform, and write data for machine learning workflows. You can interact with the SDK in any Python environment, including Jupyter Notebooks or your favorite Python IDE.

This python SDK includes the following functionality to help prepare your data for modeling:

• Intelligent time-saving transformations such as:
  • Fuzzy grouping
  • Derived column by example
  • Automatic split columns by example
  • Impute missing values
  • Automatic join
• Automatic file type detection. The SDK can automatically detect any of the supported file types. You don’t need to use special file readers for CSV, text, Excel, etc., or to specify delimiter, header, or encoding parameters.
• Cross-platform functionality with a single code artifact. The SDK also allows for dataflow objects to be serialized and opened in any Python environment. Write to a single SDK and run it on Windows, macOS, Linux, or Spark in a scale-up or scale-out manner. When running in scale-up, the engine attempts to utilize all hardware threads available, when running scale-out the engine allows the distributed scheduler to optimize execution.
• Summary statistics can be generated quickly for a dataflow with a single line of code.
• Scale through streaming. Instead of loading all the data into memory, the SDK engine streams data for better scale and performance on large datasets.

Install the SDK

To install the Azure Machine Learning Data Prep SDK for Python, use the following command:

pip install --upgrade azureml-dataprep

Learn how to use it

Here are some resources to help you learn more about the Azure Machine Learning Data Prep SDK for Python:

• See the Azure Machine Learning service documentation to learn how to load, transform, and write data with the SDK.
• Review a getting started notebook in GitHub of data preparation using the SDK.
• Use the API Reference to look up classes and modules.