Calling Compute with Python

By the end of this guide, you should have all the tools installed necessary for using the Rhino Compute through Python.

For an getting started video tutorial from Junichiro Horikawa with Python

This guide presumes you have Python installed on the platform:

  • Python 2.7 - Windows (32 and 64 bit)
  • Python 3.7 - Windows (32 and 64 bit)
  • Python 2.7 - OSX (installed through homebrew)
  • Python 3.7 - OSX (installed through homebrew)
  • Linux and other python versions are supported through source distributions on PyPi

Setting up a Compute Project in Python

There are a few client side tools which need to be installed in Python that are essential to communicate with the Compute server. These include:

  • - This is the part of the Rhino3dm libraries. It is a Python wrapper for OpenNurbs which contains the functions to read and write Rhino Geometry Objects. This is available as a Pip package.

    pip install rhino3dm

  • - This is a work in progress package which is meant to add classes available in RhinoCommon, but not available through Compute-rhino3d makes calls into the McNeel Cloud Compute server for these functions. It handles all the transaction authorizations and JSON data conversion.

    pip install compute-rhino3d

The first use of Compute

An example of using Python to access compute can be found in the example.

Please note that the compute Python calls are made through the package using a _ (underbar) and not a - (dash).

Next Steps

Congratulations! You have the tools to use Rhino Compute server. Now what?

  1. To see the transactional nature of Compute, read through
  2. See a list of the 2400+ API calls available for
  3. Download the Compute Samples repo from GitHub.
  4. The libraries are still very new and changing rapidly. Give them a try or get involved. Ask any questions or share what you are working on the Compute Discussion Forum