This is a public repository for the API, tools, and packages available for Datagrok™, a next-generation web-based data analytics platform. The platform is very extensible, and almost anything could be implemented as a package:
- Support for scientific domains, such as cheminformatics
- Applications, such as Clinical Case or Peptides
- Connectors to OpenAPI web services
- Visualizations, such as Leaflet
- Importing and previewing files, such as SQLite, PDF, or CIF
- Scientific methods implemented in R, Python, or Julia
- File metadata extractors, such as Tika
- Custom predictive models that work with the built-in predictive modeling , such as TensorFlow.js
- Platform enhancements, such as PowerPack or UsageAnalysis
- ... and other types of extensions documented here.
These open-source packages are free to use by anyone, although for the public environment there are some restrictions related to the server computational capacities. Organizations that deploy Datagrok on their premises also can access public packages. In addition to that, enterprises typically establish their own private repositories that contain proprietary extensions.
For developers: check out getting started and contributor's guide.
Datagrok grants free license to academic institutions to use it in any context, either research or educational. Moreover, publishing scientific methods as Datagrok packages provides a number of unique benefits that are specifically important to academia:
- Reproducible and scalable computations
- Making your research globally available by using data augmentation capabilities. The platform proactively suggests contextual actions and enriches the current object using functions implemented in R, Python, Julia, Matlab, or other language. In other words, Datagrok not only can run a function, but also suggests what could be derived from your dataset. This cross-pollination of knowledge could be transformative within and across a broad range of scientific disciplines.
For academic collaborations, please email [email protected].
If you want to get familiar with the platform, here are some ideas. Pick whatever interests you, and reach out to Andrew ([email protected]) or post on our community forum.
- Visualizations
- Gantt chart
- Port visjs-based network diagram from Dart to JavaScript
- WebGL-based rendering of the 2D scatter plot to work with 10M+ points
- Event drops
- Scientific methods
- Statistical hypothesis testing
- Bayesian statistics
- Computer vision
- NLP
- File editors and viewers
- File metadata extractors (see Apache Tika)
- WASM-based support for digital signal processing
- Domain-specific algorithms
- Connectors to web services and open datasets
- Bioinformatics
- Telecom
- Fintech