tablecloth VS next-jdbc

Another, and definitely better for serious projects, approach is to use https://github.com/scicloj/tablecloth or things that it mentions - tech.ml.dataset and dtype-next.

The forth meeting will be at the end of October and will be dedicated to the Tablecloth dataset manipulation library by generateme, with an intro by Ethan Miller, who is nowadays involved in developing Tablecloth.

I really like geni: it is really idiomatic in its approach to Apache Spark. There are some gaps (no UDFs), and I am not sure that the project is as active as it used to be. But I still use it and find it very nice (I do have Apache Spark background already). tablecloth is an alternative dataframe library that is being used by a lot of folks in the Clojure data science world. For that matter, you should check out scicloj, and also hang out in the data channel in zulip.

At this re:Clojure workshop (Nov. 7th), @ladymeyy taught us about Tablecloth.

In both sessions, we will practice data visualization on a real-world data problem. Among other things, we will try a new data visualization library that Ashima Panjwani is working on. We will assume basic familiarity with Clojure and with Tablecloth. Both sessions will be independent, probably overlapping in content.

We are planning some Scicloj study sessions this weekend about data wrangling with Tablecloth.

And in general for R integration and more data science goodies checkout scicloj and in the vein of dplyr style extremely thought out interfaces I highly recommend tablecloth.

I can't help you with the database part but I think you have enough grist for your mill, so I would use whatever db you know and then connect to it via next-jdbc

3: Postges is fine, i'd recommend next.jdbc to connect to postgres.