Provides a Jupyter workflow for running nanshe.
Clone or download
Pull request Compare This branch is even with nanshe-org:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

Build Status License Release Gitter

Nanshe Workflow


Provides a Jupyter workflow for running nanshe.


There are several requirements that must be satisfied to run the workflow. Also there are some optional requirements for some extensions or better performance. A list of the direct requirements can be found in the included recipe. Basically all of these dependencies are available from conda-forge with the exception of nanshe, which is available from the nanshe channel.


The preferred method of installation is to use Docker as explained below. This supports all major OSes and is the easiest way to get started. In some cases, this might not be possible or desirable, in which case conda can be used to install natively.


The easiest way to get started with the workflow is to use docker. It is available for all major OSes and is easy to install. Once installed, we provide directions for getting the container up and running that have very simple requirements.


If one would rather run the workflow natively instead of using Docker, we provide an alternative installation method for Mac and Linux. Windows is currently not supported. Simply install Miniconda or Anaconda based on your preference. Nearly all of the dependencies are available from the conda-forge channel. To add it simply run conda config --add channels conda-forge. The rest of the dependencies are in the nanshe channel. It can be added in the same manner conda config --add channels nanshe. Once done simply use conda-build to build the included metapackage by running conda build nanshe_workflow.recipe. Once this is complete install into your current environment by running conda install --use-local nanshe_workflow and you should be ready to go.


These instructions are primarily for usage with conda. For typical usage with the Docker container, please see these instructions.


To start up the workflow, simply open a terminal and run jupyter notebook nanshe_ipython.ipynb. This will open a new tab/window in your browser with the workflow loaded.


Each cell with user input has an explanation beforehand with the relevant parameters noted and how they can be set. The second cell should be used to provide input. Later cells are used to specify various relevant parameters. To run a cell, just press the key combination Shift + Enter. This will take you to the next cell as well. Some cells may also show interactive GUIs so as to explore the result from each operation.


If you intend to work on the workflow, then you must install jq and have it on your path. A package for it exists in conda-forge. After cloning the repo, make sure to include the .gitconfig file in your .git/config. If just cloning the workflow, this should be as simple as running echo -e "[include]\n\tpath = ../.gitconfig" >> .git/config. This only needs to be done once.