Basic support for non-string names.

[ueshin]

Currently names in Koalas, e.g., df.columns, df.colums.names, df.index.names, need to be string or tuple of string, but it should allow other data types which are supported by Spark.

before:

>>> kdf = ks.DataFrame([[1, 'x'], [2, 'y'], [3, 'z']])
>>> kdf.columns
Index(['0', '1'], dtype='object')

after:

>>> kdf = ks.DataFrame([[1, 'x'], [2, 'y'], [3, 'z']])
>>> kdf.columns
Int64Index([0, 1], dtype='int64')

[itholic]

Overall it looks fine, but I wonder maybe if it wouldn't matter to abandon some of the features we could have supported before ??

[itholic]

I think maybe we don't need to consider tuple here because Index cannot have tuple value in Koalas anyway ??

>>> ks.Index([(1, 2), 3])  # Cannot have a tuple for Index as mixed type
Traceback (most recent call last):
...
pyarrow.lib.ArrowInvalid: Could not convert (1, 2) with type tuple: did not recognize Python value type when inferring an Arrow data type

>>> ks.Index([(1, 2), (3, 4)])  # This is not gonna be an Index but MultiIndex
MultiIndex([(1, 2),
            (3, 4)],
           )

Oh, but if we had any plan for supporting tuple for values, I think It's okay to remain the TODO as it is.

[ueshin]

Let me revert some commits. Maybe we should discuss input checks separately.

[itholic]

Looks good enough in terms of Basic support !

Let's discuss the details separately.

[ueshin]

Thanks! I'd merge this now. I'll submit PRs for each input check.