Rewrite dict literal for files in pandas/tests/frame/test_indexing and test_where

[Qbiwan]

• closes CLN: C408 Unnecessary dict call - rewrite as a literal #38138
• tests added / passed
• passes black pandas
• passes git diff upstream/master -u -- "*.py" | flake8 --diff
• whatsnew entry
  identify unnecessary dict call - rewrite as a literal
  Rewrites as dictionary literals for the following files in two spots:
• pandas/tests/frame/test_indexing
• pandas/tests/frame/test_where
  based on @jreback comments:
  Rewrite dict literal for files in pandas/tests/frame #38207 (comment)

[Qbiwan]

@jreback checks showing some issues: pandas/tests/frame/indexing/test_indexing.py:77: error: "Set[Iterator[Tuple[Any, int]]]" has no attribute "keys" [attr-defined]

maybe its because replacing dict() with {} in this situation would cause python to mistake {} to be set(). Should we leave it as dict() for this case?

l=[1,2,3]
a = dict(zip(l, range(len(l))))
type(a)
>> dict

b = {zip(l, range(len(l)))}; 
type(b)
>> set

The second change, which is for pandas/tests/frame/test_where, might also show a problem as using dict() gives us a dictionary of three items correctly

df = {1:2,2:3,3:4}
dict((c, s + 1)  for c, s in df.items())
> {1: 3, 2: 4, 3: 5}

but using {} gives us a set of 3 tuples, which might not be what we intended

{(c, s + 1)  for c, s in df.items()}
> {(1, 3), (2, 4), (3, 5)}

[fangchenli]

You could use dict comprehensions.

l = [1,2,3]
x = {value: index for index, value in enumerate(l)}
x 
>> {1: 0, 2: 1, 3: 2}