Wakari provides a powerful set of pre-configured Python environments, but what if you need a Python module that Wakari doesn’t include by default? No problem, just login to your Wakari account, open a shell terminal and install the module you need. It is really that simple.
What Are My Options?
Because Wakari is built on Anaconda you have several options for installing additional Python modules.
- conda pip
conda is the package management tool that comes with Anaconda. With conda you can install new modules or upgrade existing modules that are part of the Anaconda package library. However, conda also supports a larger ecosystem of third-party repositories that can also contain packages not included in Anaconda by default. See the blog post on conda for more details.
But I Use pip or easy_install
Not a problem. You can use either pip or easy_install on the command line in Wakari just like you would on your local system.
Let’s say you want to install the pyephem package so you can locate the position of a comet or a planet.
In addition, you may uninstall with pip as well.
Reset an Environment
Often times experimentation means starting over. If, after customizing an environment, you decide you would like to reset things back to their original state, you can easily reset the environment using the Reset operation on the PyEnvs tab. Simply chose the environment from the envs list, chose Reset from the operations list, and click on the Go button. All changes made to the environment will be removed and the environment will be restored to its original, default Wakari state.
Create an Environment
Using the conda command it is also possible to create entirely new environments with a completely unique set of packages. This is especially desirable if you intend to share the environment, and you want to provide an environment which only includes the required set of packages to enable your simulation or analysis project.
When conda creates a new environment or installs additional packages, it will also include all required package dependencies.
Here is an example of creating a new environment called “mysimenv” that includes the popular matplotlib package and all its dependencies.
Share an Environment
The Wakari team is currently building the feature of sharing an environment with others. Once you have set up an environment and built out your analysis, it will only be helpful to others if they can run it too. With “shareable environments”, you will be able to tell Wakari to snapshot your environment, so that you can share with others. They will then be able to recreate the environment within their Wakari workspace and run your code without any hassles.
Custom environments have opened up a whole new set of possibilities within Wakari. Whether you use Wakari in the cloud or are looking to install Wakari within your company’s private cluster, custom environments make it easier than ever to collaborate and reproduce data analysis with Python.
Give it a try and let us know what else would be cool in Wakari.