# @datagrok-libraries/arrow > DataFrame ↔ Arrow IPC conversion (extracted from @datagrok/arrow) Latest version **1.1.2** (published 2026-09-10) · 0 weekly downloads ## Install ```sh npm install @datagrok-libraries/arrow pnpm add @datagrok-libraries/arrow yarn add @datagrok-libraries/arrow bun add @datagrok-libraries/arrow ``` ## Health **Score 65/100 (B)** — status: active. Positive: has types; no vulnerabilities; has provenance; recently updated; high maintenance score. Warnings: low downloads; no esm support. ## Facts | | | |---|---| | Version | 1.1.2 | | Published | 2026-09-10 | | First published | 2026-04-29 | | Weekly downloads | 0 | | TypeScript types | bundled | | Module format | CommonJS | | Dependencies | 2 | | Unpacked size | 44 KB | | Known vulnerabilities | 0 | | Install scripts | no | | Provenance | attested (GitHub Actions) | | GitHub stars | 73 | | Author | Pavlo Polovyi | | Maintainers | aparamonov_datagrok, drizhina, spodgrok | ## Links - npm: https://www.npmjs.com/package/@datagrok-libraries/arrow - Repository: https://github.com/datagrok-ai/public - Homepage: https://github.com/datagrok-ai/public#readme - Issues: https://github.com/datagrok-ai/public/issues - npm.io page: https://npm.io/package/@datagrok-libraries/arrow ## Dependencies (2) - [apache-arrow](https://npm.io/package/apache-arrow.md) ^21.0.0 - [datagrok-api](https://npm.io/package/datagrok-api.md) ^1.27.0 ## Recent versions - 1.1.2 (latest) — 2026-09-10 - 1.1.1 — 2026-08-31 - 1.1.0 — 2026-06-03 - 1.0.0 — 2026-04-29 ## README # Datagrok package repository This is a public repository for the API, tools, and [packages](https://datagrok.ai/help/develop/develop#packages) available for [Datagrok™](https://datagrok.ai), 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](packages/Chem/README.md) * Applications, such as [Clinical Case](packages/ClinicalCase/README.md) or [Peptides](packages/Peptides/README.md) * Connectors to [OpenAPI web services](https://github.com/datagrok-ai/public/tree/master/packages/Swaggers) * Visualizations, such as [Leaflet](packages/Leaflet/README.md) * Importing and previewing files, such as [SQLite](packages/SQLite), [PDF](packages/FileEditors/README.md), or [CIF](packages/NglViewer/README.md) * Scientific methods implemented in R, Python, or Julia * File metadata extractors, such as [Tika](packages/Tika/README.md) * Custom predictive models that work with the built-in [predictive modeling](help/learn/predictive-modeling.md) , such as [TensorFlow.js](packages/TensorFlow.js/README.md) * Platform enhancements, such as [PowerPack](packages/PowerPack/README.md) or [UsageAnalysis](packages/UsageAnalysis) * ... and other types of extensions documented [here](help/develop/packages/extensions.md). These open-source packages are free to use by anyone, although for the [public environment](https://public.datagrok.ai) there are some restrictions related to the server computational capacities. Organizations that deploy Datagrok [on their premises](help/develop/admin/architecture.md#deployment) 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](help/develop/develop.md) and [contributor's guide](CONTRIB.md). ## Building The repository is one pnpm workspace with a Turborepo task graph: `grok setup` once (or `pnpm install` at the root), then `grok build` in any package (or `grok build --all`). Details, and what a package declares, are in [packages/BUILD.MD](packages/BUILD.MD). ## Academia 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](help/compute/compute.md) * Making your research globally available by using [data augmentation](help/discover/data-augmentation.md) capabilities. The platform proactively suggests contextual actions and enriches the current object using [functions](help/datagrok/functions/function.md) implemented in [R, Python, Julia, Matlab, or other language](help/compute/scripting/scripting/scripting.mdxx). 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 `info@datagrok.ai`. ## Ideas for contributions If you want to get familiar with the platform, here are some ideas. Pick whatever interests you, and reach out to Andrew (askalkin@datagrok.ai) or post on our [community forum](https://community.datagrok.ai/). * Visualizations * Gantt chart * Port visjs-based [network diagram](https://datagrok.ai/help/visualize/viewers/network-diagram) from Dart to JavaScript * WebGL-based rendering of the 2D scatter plot to work with 10M+ points * [Event drops](https://github.com/marmelab/EventDrops) * Scientific methods * Statistical hypothesis testing * Bayesian statistics * Computer vision * [NLP](packages/NLP) * File editors and viewers * File metadata extractors (see Apache Tika) * [WASM-based support for digital signal processing](packages/DSP) * Domain-specific algorithms * Connectors to web services and open datasets * [Bioinformatics](packages/Bio) * Telecom * Fintech ## See also * [Datagrok home](https://datagrok.ai/) * [JavaScript development](https://datagrok.ai/help/develop) * [Community forum](https://community.datagrok.ai/) --- _Source: https://npm.io/package/@datagrok-libraries/arrow · Machine-readable twin of the npm.io package page. Health data is recomputed on every publish._