Michel Lang

The JSON API partly provides values as strings, although they probably should be numbers. Some examples: * https://www.openml.org/api/v1/json/data/15 -> `id`, `file_id` (also note that `did` and `file_id` are integers during...

I've started, but there are still some lines handling multifid.

Set environment variable _R_CHECK_TESTS_NLINES_

12

See R internals. Otherwise the test output is truncated to the last 13 lines.

Requested via private mail and also here for mlr: https://stackoverflow.com/questions/60110684/is-there-a-way-to-set-the-base-margin-for-xgboost-in-the-mlr-framework This could just be a `p_fct()` specifying the right column.

These options should be set to get warnings for (unwanted) partial argument matching: ```r options( warnPartialMatchDollar = TRUE, warnPartialMatchArgs = TRUE, warnPartialMatchAttr = TRUE ) ``` --> https://github.com/mlr-org/mlr3/blob/main/tests/testthat/setup.R and https://github.com/mlr-org/mlr3/blob/main/tests/testthat/teardown.R...

split/apply/combine paradigm

6

I'd like to get started on this one and use this tracker to collect and discuss ideas. AFAIR @lawremi suggested back in September to use `split`/`by` (split), `bp*apply` (apply) and...

Re-upload to CRAN

5

The package has been archived. Will there be another attempt to get it on CRAN?

Logloss

2

LogLoss is not defined for p=0 and p=1. Other toolkits clip to [0+eps, 1-eps] to overcome this: https://scikit-learn.org/stable/modules/generated/sklearn.metrics.log_loss.html