Albert Y. Kim

icon indicating a website link http://rudeboybert.rbind.io/

icon company @SmithCollege-SDS icon location Northampton MA, USA icon location Assistant Professor of Statistical & Data Sciences, Smith College

Results 29 comments of Albert Y. Kim

Have `View()` function that provides a more informative error

While I agree this would be nice to have, IMO the juice isn't worth the squeeze of figuring out how to program it

Is there a problem with geom_parallel_slopes?

Hi Roger, thanks for posting this. We can look into this. I'm tagging @echasnovski who wrote this function, however I'm not entirely sure how responsive he can be these days.

Function for 3D scatterplots with regression planes

Based it on this code on [GitHub Gist](https://gist.github.com/rudeboybert/9905f44013c18d6add279cf13ab8e398)

Extend get_regression_X() functions to glm objects

Punted to Fall 2019 when @rudeboybert teaches multiple regression

Allow for more than one RHS variable in get_correlation()

Hmm, not sure about the tidy data frame tho, as I feel it deviates too much from the much more standard correlation matrix format, which IMO is easier to read....

Allow for more than one RHS variable in get_correlation()

Also throwing an error message if one of the RHS variables is categorical `fct` or `chr` "Correlation coefficients only exist between numerical variables."

Allow for more than one RHS variable in get_correlation()

From @ismayc: Looks like we might be able to use the `pairwise_cor()` function in {widyr} to do pairwise correlations if you want: https://cran.r-project.org/web/packages/widyr/vignettes/intro.html. We can even create a wrapper so...

Allow for more than one RHS variable in get_correlation()

@ismayc How about we punt on this one?

Fix digits rounding

This roots in how tibbles are "opinionatedly" printed in the console as described [here](https://community.rstudio.com/t/why-are-some-digits-in-the-tibble-output-faded-out-in-the-r-console-display/4330/19). tl;dr: The first 3 significant figures/digits represent >99.9% of the value of a number, so all...

Create `rep_slice_sample(n)`

We currently import `rep_sample_n()` from infer. The development branch of infer has the update, but it's not on CRAN yet: https://github.com/tidymodels/infer/pull/325. You ok with waiting until this is updated?