Current state of play re DataFrames, CategoricalArrays, NullableArrays, CSV and all that

[Douglas Bates]

I am starting on a project that will involve creating a data/modeling pipeline.  The data originate in a CSV file.  I would prefer to start the project using CategoricalArrays and NullableArrrays if the switch of DataFrames to default to these types is to occur soon.  Also, the release version of the CSV package returns NullableArrays by default.

Is there anything special that I need to do to work with NullableArrays and CategoricalArrays within DataFrames?  Should I check out the nl/nullable branch?  Last time things got messy.

[Milan Bouchet-Valat]

Nothing special AFAIK. The branch passes the tests, so it shouldn't
require more checking it out. Though I'll have to update it soon for
the new CategoricalArrays release (to be registered soon).

Please report any bugs you find, that would be very helpful.


Regards