Github user sethah commented on the issue:
https://github.com/apache/spark/pull/16194
One problem I see is that some models have relevant information stored in
class members instead of params. LogisticRegression for example would probably
list the number of classes and the number of features, neither of which are
params. Also, as @zhengruifeng mentioned, there are a lot of not-useful params
that we would end up listing. However, I do not like the idea of implementing
something different for every single estimator/model either.
I guess I'm not entirely convinced the added benefit is worth the cost.
What is the motivation for this, other than that it is convenient in the repl?
Are there users asking for it?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
