OpenPathSampling currently only works on Mac and Linux. It is tested against Python 2.7, 3.6, and 3.7, although there may be some corners of the code that aren’t Python 3-compatible yet.


As of OpenPathSampling 1.1, OpenMM will no longer be automatically installed when you install OPS. However, the OpenMM engine will be immediately available if you install OpenMM yourself. See the OpenMM installation instructions for a detailed guide, but conda install -c conda-forge -c omnia openmm will work for most people. (COMING SOON: details on OPS integrations with other tools.)

Standard Install with Conda

We recommend using conda to install OpenPathSampling. conda is a Python package manager built for scientific Python, and which handles binary dependencies seamlessly. If you don’t want the full conda installation, the miniconda package provides much of the convenience of conda with a smaller footprint.

OpenPathSampling is part of the conda-forge channel in conda. To install the most recent release of OpenPathSampling with conda, use the following command

$ conda install -c conda-forge openpathsampling

With that, you should be ready to use OPS!

Developer Install with Conda

To install a developer version of OPS (using conda), change to a directory where you want to OPS code (i.e., if you want the OPS git repository at directory/openpathsampling/, change to directory/.) Then download the and run it

$ curl -OLk
$ bash

At this point, any changes to the code in that download of the OPS directory will be live in your Python installation. You can use experimental code from other forks by adding the fork as a remote and checking out a branch. You can combine changes from multiple users by merging them into a branch in your local version of the repository.

Additional functionality of the script is described in Quick bugfix/developer installation.

Manual Installation

If you don’t want to use conda, you will have to manually obtain the dependencies, which you can see listed under install_requires in Then you can install from our GitHub repository.

Clone the source code repository from GitHub:

$ git clone

Then, in the directory containing the source code, you can install it with

$ python install

Or, for a developer install

$ python develop

Testing Your Installation

Running the tests is a great way to verify that everything is working. The test suite uses pytest and, for legacy reasons, also requires the nose package. You can pick these up via conda if you don’t already have them.

$ conda install pytest nose

From the source directory openpathsampling/tests, you can run the tests by typing py.test on the command line. The test suite includes over 900 individual tests, and runs in around 2-3 minutes.

Beta testing experimental updates

In rare cases, you may want to test code that hasn’t been merged into the core of OPS yet. Instructions to install in this case are in the docs for Installing for developers and beta-testers.