Support for times being a pd.Timestamp

[AdamRJensen]

Right now the times parameter has to be a pd.DatetimeIndex. While a single timestamp can be passed, this is kinda awkward as you would normally just pass a timestamp as ``pd.Timestamp`.

The function _pandas_to_utc could be modified to allow for a single time stamp being passed to the solar position algorithms. Is this a desirable feature?

The functionality would look like this:

solposx.solposition.michalsky(
    times=pd.Timestamp('2020-01-01 00:00+00'),
    latitude=50,
    longitude=10)

alternatively maybe even with a single timestamp in the form of a string:

solposx.solposition.michalsky(
    times='2020-01-01 00:00+00',
    latitude=50,
    longitude=10)

This could be achieved with addition of the following line to the _pandas_to_utc function:

    if isinstance(pd_object, (str, pd.Timestamp)):
        pd_object = pd.DatetimeIndex[pd_object]

[kandersolar]

Would the returned solar position angles still be in a DataFrame? A Series (or dict) might be more consistent, but also perhaps more surprising...