sas-cvpy 1.1.0

SAS Computer Vision Interface

CVPy - SAS Viya Computer Vision API for Python

Overview

CVPy is a Python package that makes SAS image analytics more accessible to Python users in a variety of different ways. This high-level Python library provides helpful APIs that assist with the processing, analyzing, and visualization of images. This allows users of the image and biomedimage action set in SAS Viya to have more flexibilty when working in Python.

Currently available within CVPy are visualization APIs that allow users to easily be able to move images out of CAS and into common open source tools such as Mayavi and Matplotlib. These APIs streamline the visualization of image data fetched from a CAS table and can assist with further image analysis.

Prerequisites

• Python version 3 or greater is required
• Install SAS Scripting Wrapper for Analytics Transfer (SWAT) for Python using pip install swat or conda install -c sas-institute swat
• Access to a SAS Viya 3.5 environment with Visual Data Mining and Machine Learning (VDMML) is required
• A user login to your SAS Viya back-end is required. See your system administrator for details if you do not have a SAS Viya account.
• Install Mayavi for scientific data visualization in Python

Mayavi Installation

Mayavi is a heavy, complex package that is required by CVPy. If you do not already have Mayavi installed, follow the steps below to install it.

Install VTK, a critical dependency for Mayavi.

pip install VTK‑8.1.2‑cp37‑cp37m‑win_amd64.whl

or

pip install VTK==8.1.2

Note that a specific version of is VTK used. You can download the wheel file from here.

Install Mayavi

pip install mayavi

You may see an error message about numpy, like “Numpy is required to build Mayavi correctly, please install it first.” Ignore that message.

Install PyQt, a GUI toolkit needed to run Mayavi

pip install PyQt5

CVPy Installation

To install CVPy, use the following command:

pip install sas-cvpy

Contributing

We welcome your contributions! Please read CONTRIBUTING.md for details on how to submit contributions to this project.

License

This project is licensed under the Apache 2.0 License.

Additional Resources

• Python-CVPy API Documentation
• Biomedimage action set
• Image action set