Datalore Improvements in 2020: Datalore Professional, a Better Coding and UI Experience, and More
Read this post in other languages:
Last year presented an unexpected challenge for the world, one we all had to rise to. In 2020, we managed to launch Datalore Professional and deliver lots of great new features for our users. Here we’ve assembled a description of some of the most notable ones.
In November we launched Datalore Professional. It was designed for solving more complex tasks with larger datasets that need more powerful computation hardware.
Here is a comparison table for Datalore Community and Datalore Professional. Take a closer look, and decide which plan best fits your needs!
|Basic machine (4 GB RAM, AWS name: t3.medium)||120 hours||♾️|
|Large machine (16 GB RAM, 2 vCPU cores, 400% faster than the basic machine, AWS name: r5.large)||–||120 hours|
|GPU machine (1 NVIDIA T4 GPU, 16 GB GPU RAM, 4 vCPU cores, AWS name: g4dn.xlarge)||–||20 hours|
Better coding experience
Code insight from PyCharm
At the beginning of the year we integrated code insight from PyCharm, bringing autocompletion, refactorings, quick-fixes and navigation into Datalore. Now you can enjoy the same top-level PyCharm coding assistance in your online Jupyter notebooks.
Jupyter kernel compatibility
Last year we also worked hard on the stability of the IPython kernel in Datalore, and we can now say that the kernel is a lot more reliable. There is still some room for improvement, and we’ll continue working on it this year.
The Jupyter kernel is now fully supported, with widgets, plotting libraries, and shell commands.
Notebook startup times have also been improved. Notebooks now load significantly faster, as we’ve managed to cut the time it takes Conda to start up in half!
Have you ever heard of Kotlin? It is an open-source programming language developed by JetBrains. You can use Kotlin to do data science and develop multiplatform applications. In Datalore we’ve added support for Kotlin in IPython notebooks. Give it a try! Just choose Kotlin as your language when creating your notebook.
Workspace files and S3 buckets support
This December we implemented support for Workspace files, making it possible for you to share the data across several notebooks.
For those of you who work with lots of data, we also added support for mounting S3 buckets. Learn more about it from this blog post.
Better UI experience
Sidebar for quick actions
To help you work faster with data files and navigate the contents of notebooks more easily, we added a sidebar tab inside the editor. With the sidebar you have direct access to attached files, including notebook and workspace files. You can also use the Table of contents and Variable viewer there. The Shortcuts window will also appear in the sidebar when opened from the Help menu.
Last year we also introduced dark mode for Datalore. You can change the theme of the notebook in the View menu tab inside the editor, where you can also enable Distraction free mode and the Split view options.
Last year we introduced a toolbar for better Markdown editing. It helps you describe your code with text, LaTex formulas, and HTML code inside the Markdown cells.
JetBrains has a long history of collaboration with Anaconda, and PyCharm is a Python IDE recommended in the Anaconda installer. As of October 2020, Datalore and PyCharm are both featured in the new Anaconda Navigator! Update your Anaconda Navigator to the latest version and launch Datalore directly from it.
Tutorials and research
Last year we did some interesting research and tutorial projects:
- We Downloaded 10,000,000 Jupyter Notebooks From Github and Shared the Dataset with Everyone
- We Analyzed 495 AMD Radeon and Nvidia GPU Specifications and Shared the Dataset with Everyone
- Getting Started Video Tutorial
- Visualization with Pyplot Tutorial
- Advanced Visualization Tutorial on a Fresh GPU Model Dataset
Check them out!
That’s it for Datalore’s major updates from 2020. Follow our blog for information about the new updates being introduced this year.
The Datalore team