DOC: reverse rolling window (#38627) (#41842)

* DOC: Add examples to do reversed rolling window (#38627).

File is copied from @vangorade last changes in pull request (#39091):
#39091

Fixed typo to pass pre-commit codespell hook.

* DOC: Add some suggested changes.

From pull request (#39091):
#39091

- remove noqa
- change line back to single quotes
- rename df1 to reversed_df

* DOC: Add reference to FixedForwardWindowIndexer.

* DOC: Add more suggested changes.

From pull request (#39091):
#39091

- remove repeated df initialization in code example
- rename df2 to reversed_df

* DOC: Add suggested changes to reverse rolling window code examples (#38627).

Move code example as suggested in pull request (#39091):
#39091

Compiled document to check it looked fine by running:
python make.py --single user_guide/window.rst

* DOC: Revert more double quotes to single quotes.

Unecessary changes.

* DOC: Change binary window description to match master branch.

Description was not modified after fetching and merging, so pasted from master branch:
https://github.com/pandas-dev/pandas/blob/master/doc/source/user_guide/window.rst

* DOC: Change exponentially weighted mean formula to one in master branch.

Since fetching and merging didn't modify this line.

* DOC: Replace window.rst with file from pandas-dev master branch.

To only add reverse rolling window section and avoid any unwanted changes.

* DOC: Add section with reverse rolling window example.

* DOC: Add section markup.

Forgot to add.

* DOC: Delete the same code example.

FixedForwaredWindowIndexer example is same as the one above the added section.

* DOC: Remove section and first sentence.

Modify sentence slightly to show this example should output same result.

* DOC: Remove unecessary words.

Coding is already in Python.

Author: mdhsieh

doc/source/user_guide/window.rst

@@ -299,6 +299,24 @@ forward-looking rolling window, and we can use it as follows:
    indexer = FixedForwardWindowIndexer(window_size=2)
    df.rolling(indexer, min_periods=1).sum()
 
+We can also achieve this by using slicing, applying rolling aggregation, and then flipping the result as shown in example below:
+
+.. ipython:: python
+
+   df = pd.DataFrame(
+       data=[
+           [pd.Timestamp("2018-01-01 00:00:00"), 100],
+           [pd.Timestamp("2018-01-01 00:00:01"), 101],
+           [pd.Timestamp("2018-01-01 00:00:03"), 103],
+           [pd.Timestamp("2018-01-01 00:00:04"), 111],
+       ],
+       columns=["time", "value"],
+   ).set_index("time")
+   df
+
+   reversed_df = df[::-1].rolling("2s").sum()[::-1]
+   reversed_df
+
 .. _window.rolling_apply:
 
 Rolling apply