JetBrains recently introduced Datalore 1.0, an intelligent web application for data analysis and visualization, which is focused specifically on the machine learning environment in Python. This version brings an improved smart code editor, user-controlled code execution, a professional subscription option, and more.
Datalore smart code editor aims to make coding for machine learning enjoyable and easy, providing features such as autocompletion, syntax highlighting, inspections, quick-fixes, easy navigation, intentions (context-dependent suggestions in input blocks).
(image taken from blog.jetbrains.com)
Datalore comes with a built-in version control system to help users manage modifications in the code; this feature provides a way for users to return to an earlier model and recover previous edits, add checkpoints for future reference, track inactivity savepoints and fix issues easily. Furthermore, it is possible to share a workbook with colleagues providing real-time access to workbook and code editor.
Users can now choose two possible ways to run code in the application. The first is putting the workbook on hold to complete major code edits, running only the computations that the user wants to check right away. In the second, users can keep using online code execution that automatically runs calculations and applies changes in the code. It is important to notice that in both cases, the incremental recalculation takes care of the consistency of the workbook.
(image taken from blog.jetbrains.com)
Data analysis with Python usually necessitates numpy, pandas, sklearn, etc. Datalore comes with some out-of-the-box pre-installed Python libraries for data analysis. Users can easily import pandas, numpy, and matplotlib to toggle with dataframes, build Pytorch-powered neural networks, etc. In addition, Datalore brings two libraries for advanced visualization, which are:
- datalore.plot, inspired by the "grammar of graphics" ideas and their R implementation ggplot.
- datalore.geo_maps, which enables the addition of interactive maps for analysis.
JetBrains provided two plans for users: the community plan which offers 10GB of upload space and up to three computations running in parallel, and the professional plan subscription for users who require extended data storage and high performance computational resources, which offers 50GB of upload space and up to 10 computations running in parallel. Professional plan subscription starts with $20 monthly.
More information can be found on datalore.io.