mb706

That one would be useful, but an alternative that combines the strings (as `assert()` does) could be nice as well. I.e. ```r `%check&&%` = function(lhs, rhs) { if (!isTRUE(lhs) &&...

We can discuss this, maybe we even want https://github.com/mlr-org/mlr3pipelines/issues/572

Official stance of mlr3 is now that we should solve this "manually" from within mlr3pipelines.

This should probably be dependent on a parameter, however.

problem to anticipate here is that not all resamplings create a prediction for every input row (https://github.com/mlr-org/mlr3pipelines/issues/216 will get relevant here), and some may create multiple predictios (does mean averaging...

filling with `NA` is probably a good idea. I don't like multiple columns because the number of output columns must be the same in train() and predict. I guess if...

I'm not sure how `mlr3` currently handles / should handle `Prediction` with missing rows. We may also want to use classifavg / regavg for unioning because they would support multiple...

Also the need to filter rows during training (e.g. based on NA or some other property that a learner cannot handle) may be very different from this splitting up predictions...

That argument for 'nacols" would also apply to other conditions to filter by, e.g. trying to filter if any of a number of features is 0 or negative. We could...

So this is supposed to filter out rows during predict time, right? Is there a way to keep track of what sample is being predicted, once a learner sees the...