[SPARK-11889] [SQL] Fix type inference for GroupedDataset.agg in REPL

[marmbrus]

In this PR I delete a method that breaks type inference for aggregators (only in the REPL)

The error when this method is present is:

<console>:38: error: missing parameter type for expanded function ((x$2) => x$2._2)
              ds.groupBy(_._1).agg(sum(_._2), sum(_._3)).collect()

[marmbrus]

@dragos any idea why it fails to infer the type only in the REPL?

[dragos]

I can't reproduce the difference. It won't infer it in a standalone program either.

As I mentioned in our conversation, it's a chicken and egg problem: type inference is guided by the expected type, but if the method is overloaded, the expected type is not known. And the argument type is what guides overload resolution. It works in simple cases, when the overloads have different aritites, but with varargs that's no longer the case.

[marmbrus]

It always works fine here though. Why is this different?

[SparkQA]

Test build #46441 has finished for PR 9870 at commit d7367da.

• This patch fails Spark unit tests.
• This patch merges cleanly.
• This patch adds no public classes.

[marmbrus]

Ugh, sorry :(

I'll fix this.