Yesterday while trying to launch a new Jupyter notebook instance from Anaconda Navigator. JupyterLab is a new interface for python development from the house of Jupyter which is still under alpha development. For users who are familiar with Jupyter, JupyterLab is pretty easy. In fact, JupyterLab is a space to consolidate a notebook, console, text file etc. in one single UI.
You can install JupyterLab from within Anaconda. Once installed it will appear in the list of Applications.
The UI opens to the Launcher UI. This tab will display all the environments available in the Anaconda instance. This could include Julia or R environments if configured. Apart from accessing the environments in console or Jupyter notebooks, we can open a bash console or a text file.
As mentioned earlier JupyterLab can consolidate all views as tabbed views.
You root directory is displayed in the sidebar. We can access all the files available. Surprisingly even in the alpha stage JupyterLab can render .jpeg images and .csv files lying in my root directory.
Managing instances and tabs
You could also pop out the tabs out of the view and place it side by side for better viewing.
The Running tab shows all the running instances. You can manage the sessions from this.
The command tab acts as a guide to using commands. This makes it a little easier than Jupyter notebooks with easy reference to commands.
The Tabs menu shows running tabs. However, there is no real application of this tab.
JupyterLab is a great UI to work with especially for data science. For a person who is used to Jupyter and Rstudio for Data science, JupyterLab makes it really easy to code in Python. It seems pretty stable in the alpha stage and will only improve towards the beta. It surely is a tool to look forward to for data science PoC.