-
First off, wonderful package! this really helped the community to workout the pre-processing of data.
I am having a bit of trouble wrapping my mind around the variable selection, say I want to sele…
-
From tidymodels/tidymodels.org#174
Ah, I think the formula has been repaired for parsnip but **not** for workflows.
Check this out:
``` r
library(tidymodels)
#> ── Attaching packages ─────…
-
Hi,
Just wondering why you have to specify `truth = ...` and `estimate = ...` for the class-based metric sets as shown below, but not for the numeric based metrics. Maybe its something obvious I'm…
-
## The problem
The `resamples` argument of `tune_grid` says it takes an `rset()` object. `rsample::apparent` returns a rset object, but it does not work in `tune_grid`.
The docs say to avoid `r…
-
Hello Arvid,
I want to ask you a question. Maybe qBitTorrent guys, can also add something about this.
Are there any gains when using RC1_2 compared to RC1_1?
Maybe lower memory/cpu usage? Bi…
-
Hi Max & Davis,
Loving the `tidymodels` suite!
I'm experimenting with the `parsnip` + `dials` interfaces, and I'm doing great until I get to the "merge". I would expect to be able to use a funct…
-
-
Hi,
Perhaps I'm missing something, but I am unable to get the example from the documentation [here](https://tidymodels.github.io/dials/articles/articles/Dials_and_Parsnip.html) to work. Reprex belo…
-
A step to convert `NA` values to "unknown" in categorical data, such as
``` r
library(tidymodels)
#> Registered S3 methods overwritten by 'ggplot2':
#> method from
#> [.quosures …
-
In the current devel version, `mode` is required for models that can fit classification or regression.
We need a better error message when it is left to be "unknown" when `fit()` or `fit_xy()` are…