hi Jonathan,
Well, one of the good things about Apache projects is that working in
private is frowned upon, so if you don't find anything in JIRA or in
the mailing list archives then the answer is most probably "no" :)
That being said I'm interested in integrations between Apache Arrow
and machine learning frameworks. Generally some serialization is
necessary at some point because machine learning frameworks operation
in many cases on (homogeneously-typed) multidimensional arrays or
sparse matrices
- Wes
On Tue, Mar 19, 2019 at 1:42 PM Jonathan Chiang wrote:
>
> Hi,
>
> I was curious if there was any concrete roadmap to build compatibility
> between these Apache frameworks.
>
> Would love to do data processing in python/R in arrow for deep learning or
> tabular data model building.
>
> My use case is model building and interoperability between R and Python users
> at Stanford Biomedical Informatics.
>
> My dream would be to have cloud based /local hybrid deep learning and reduce
> the amount serialization needed between R and Python users.
>
> Thanks,
> Jonathan
