Marc Becker

Yes I can reproduce. I am on it. Simplified reprex. ```r library(mlr3) library(mlr3pipelines) task = tsk("iris") resampling = rsmp("holdout") graph = gunion( list( po("pca") %>>% po("learner_cv", id = "featureless", lrn("classif.featureless")),...

@mohammadreza-sheykhmousa You cannot use `po("subsample")`. The resulting task has less rows than the other tasks created by `po("learner_cv")` and therefore `po("featureunion")` cannot bind them.

`PipeOpFixFactors` is no solution for the gallery post. `PipeOpFixFactors` introduces missing values which are not supported by `regr.kknn`.

Yes but we are not quite there yet. You can install the `mlr3tuning` development version from GitHub and use `TunerGridSearch` with hotstarting. The grid is sorted by the hotstart parameter...

@mb706 Like this https://github.com/mlr-org/mlr3learners/commit/7594ad7b3f5f281d4d6cef1cfd8d0119e9c52f7a ? Do you know how we can handle multi class tasks? We get for each target class a different `beta` matrix and the positions at which...

Replaced by #235

Merge with https://github.com/mlr-org/mlr3/pull/853

@mllg Please (review and) merge

Yes this is the old RSAGA blog post. I wrote it in March 2018.

@edzer Ready