Data Science Datalore News

New in Datalore: Sidebar, Startup Time Improvement, Stack Trace Navigation, Tensorflow Code Insight, and Two New tutorials

This September has been quite productive for our Datalore team. We’ve performed a lot of user interviews and implemented some long-awaited features to improve the user experience. And now we are happy to share the new updates with you!


This month we’ve introduced the sidebar! It helps you access Attached files, the Library manager, and the Table of contents in just one click.


Notebook startup-time improvements

Notebooks in Datalore now load significantly faster. We’ve managed to cut the Conda startup time in half! We hope you can now start working on your projects even faster using Datalore.

New Getting started tutorial

Do you remember the good old “Hello world” notebook in your Datalore file system? Many things have changed since it first appeared, so we’ve recently added a “Getting started” notebook for new users.

We decided not to update the old tutorial file for existing users, as some might have added valuable code to it. However, anyone can access the Getting started notebook from here.

Stack trace navigation

If you encounter any errors in your notebook, you are now able to navigate from the stack trace to the exact place in the code where an exception was thrown. Try it out and let us know what you think via the contact form in Datalore!

Visualization tutorial on Pyplot

This month we worked with the Datalore developer advocates to create a Pyplot tutorial, which explains how to perform essential visualizations for your projects.

Learn how to create line plots, scatter plots, pie charts, and histograms. Additionally, discover how to add titles, labels, legends, and tune colors, and how to work with subplots.

Performing visualizations in Python can be quite repetitive. So please feel free to take and reuse the code for this tutorial from this published Notebook.

Tensorflow code insight

This September we added the missing code insight for tensorflow.keras and some other tensorflow modules. Now code completion, parameter info, quick-fixes, and documentation work for tensorflow as they do for PyCharm.

We hope this makes it easier for you to work on your Deep Learning projects in Datalore!

Unzipping archives in Attached files

There’s no longer any need to use the unzip terminal command or zipfile library to unpack your archives in Datalore. We now support unpacking .zip, .tar, and .tar.gz archives with just one click using the Attached files menu:

As always, we are eager for your feedback. Please don’t hesitate to write to us in the comments or post on our forum.

Enjoy your data science journey,
Your Datalore team

image description