seb.
seb.
When constructing a `H2OFrame` in Python using the constructor, the `skipped_columns` parameter systematically causes a bug due to a mismatch during the call to `parse_setup` ```python In [1]: data =...
a supplement to https://github.com/h2oai/h2o-3/issues/15857, if we want to completely remove the legacy TE integration as it's not needed anymore if we accept the pipeline support.
As AutoML wants to apply transformations to the features from the original dataset, we need to encapsulate those transformations in a Pipeline model that can then be returned to the...
When importing a scipy sparse matrix to H2O using the `H2OFrame` constructor, the `column_names` param is ignored and columns are named using the default `C1, C2, ...` convention. discovered when...
**H2O version** 3.44.0 ## Actual behavior ### in Python: ```python In [33]: fr = h2o.H2OFrame([[1, 'a'], [2, 'b']]) Parse progress: |████████████████████████████████████████████████████████████████████████████████████ (done)| 100% In [34]: fr Out[34]: C1 C2 ----...
In many places in our R API, the dots parameter is immediately converted to a R list (using `list(...)`) which causes the evaluation of all the parameters passed in this...
Sub-issue of https://github.com/h2oai/h2o-3/issues/15854 As soon as the Pipeline models don't support Mojo, we should at least provide the end-user with a complete list of the transformations that were applied to...