mlr3pipelines
mlr3pipelines copied to clipboard
Dataflow Programming for Machine Learning in R
The date features PipeOp currently only supports `POSIX()` features, but it should also support `Date()` ones and just neglect the date time specific generated features.
Apparently some behaviour was only tested by [PipeOpMissInd autotests](https://github.com/mlr-org/mlr3pipelines/blob/f8e8bc02d405c1dd8e28a1ab03c8698d8110b9e2/tests/testthat/test_pipeop_missind.R#L6) because that one has `affect_columns` set by default? See why autotests don't all do this kind of check.
* input pipeop that starts timing but does nothing * wrapper learners that are connected to that by-reference * somehow use timeout by encapsulatoin? or at least refuse to start...