how to run SQL queries on tibbles?

General
,
1

I would like to improve my SQL skills a bit using ad-hoc small tibbles. How can I run SQL queries (merges, etc) that will work on my tibbles as if they were databases?

Does that make sense?

Thanks!

2

Yes, there is a package for that

2 Likes
3

thanks! but the package looks abandoned and I wonder about various SQL flavors... perhaps that's the best we can get anyway... thanks

4

You can use dplyr::left_join, dplyr::right_join, etc. But R does work differently than SQL. I use both SQL & R a lot. I would recommend installing MySQL on your computer and loading some tables to play with. You can use MySQL Workbench. It's all free.

SQL as a language is fairly straight-forward. The challenge is thinking in set-based terms and avoiding things like cursors. There are tutorials with SQL problems that will help you sharpen your SQL thinking skills too.

I would also recommend pulling data from MySQL to R to get better at what SQL should handle and what R should handle regarding datasets.

2 Likes
5

You could try https://github.com/ianmcook/tidyquery, but the SQL support is still fairly basic.

2 Likes
6

Thanks! That’s interesting but it is true that the limitations are too important

7

Go with dbplyr. I'll link some info on how to run SQL queries from it on the database (low data crunch), sort your data to then bring it into R.

-Good start here.

-If you're up for paying for a project-based class Matt Dancho's Learning Labs give you access to 3 SQL based projects (and 20+ other business analytics projects) that show you how to use dbplyr with a business goal, data set and outcome.

See Labs 21, 22, 23; cost is $19/m.
@mdancho

1 Like
8

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.