Show HN: Copilot for Data Analysis, beyond an open-source alternative to Tableau | Hacker News Reader

Show HN: Copilot for Data Analysis, beyond an open-source alternative to Tableau

github.com

61 pointsloa_observer12 comments

RATH is an open-source software for data exploration and visualization. It automates your workflow in data analysis. You can give RATH a CSV file and it will generates visualizations with insights and patterns automatically. It also has a data painter tool for more casual interaction with charts (https://youtu.be/djqePNyhz7w). It also has a lot of cool features like autovis, causal discovery, GPT integration, predictive interaction for text feature extraction, etc. Some of RATH’s components are also independent open-source project you might be interested:

  Like

+ Graphic Walker (https://github.com/Kanaries/graphic-walker): A lite embeddable component for visual analysis.

+ PyGWalker (https://github.com/Kanaries/pygwalker): turning your pandas dataframe into a Tableau-style User Interface for visual exploration.

RATH is a collection of interesting ideas that we think the next generation of data analysis software should be, so there might be many features that not well organized to be a united app. Tell me which feature you prefer and which is not. Looking forward for your ideas and advice.

Open articleView on HN

Hi, great product. Was curious often in data exploration we need to right complex SQL queries (multiple joins and operations to get the desired results) where a lot of context is required in terms of how data is stored in these databases and get some metric for visualisation.

I am curious to understand is your product able to create automated visualisation which were created by writing complex queries on Metabase or BigQuery

RATH's autovis feature is based on a data view. If your queries generates a data view (tables with columns), it can handle it.

Thanks. Love seeing some of the power of Data Description’s DataDesk as open source code.

While I don't have a use case for this right now, it hits a nice spot that I used to have trouble finding the right sized tool for. Your readme is also very good!

If someone comes from the `voyager` ecosystem, a similiar way easy-launching it can be achieved with [r0hitsharma/voyager-cli: Launch Voyager data exploration tool against a local data source.](https://github.com/r0hitsharma/voyager-cli)

I've had my eye on this for a while. Looks promising but as far as I can tell it's paid-only? Not mentioned anywhere on the repo but on the marketing site it doesn't appear to be possible to check it out without subscribing. Which is a hard pass from me, for one.

EDIT: I see you have binaries available from the repo for the July 2022 version, but not for the most recent 2.0.0 release? Is this just an oversight perhaps?

Try this: https://rath.kanaries.net. (demo link from readme)

I believe it would be a good idea to integrate Rath's automated analysis capabilities into PyGWalker. I need it so much!

There will be a new SDK for integrate it soon.

This work looks super promising. I am excited to test it out, thank you!

Looking forward to your feedbacks! Here is a demo link you can try without login: https://rath.kanaries.net

Where exactly would our data be going once we upload it?

It will not be uploaded to Kanaries' server if you are using the open-source version of RATH. The data is stored in the indexedDB in your browser. For the SaaS version of RATH, the data will be stored in AWS S3 in the US.

Reply on news.ycombinator.com