please add quick documentation on pandas timestamp values of month, day of week

[jason-s]

Pandas appears to use values for timestamp fields where 1 = January, 2 = February, etc., (vs. 0 = January) and 0 = Monday, 1 = Tuesday, 2 = Wednesday, 3 = Thursday, 4 = Friday, 5 = Saturday, 6 = Sunday. (vs. weekday numbering that starts with 1 instead of 0, or Sunday instead of Monday.)

dec31 = pd.Timestamp('2013-12-30 00:00:00')
dec31.weekday()

prints 0 in my system (Mac OSX + pandas 0.12.0)

This convention doesn't appear to be in the documentation + it would really help as it is one of several possibilities.

[jreback]

would surely take a PRto document

these are the python conventions in any event

http://docs.python.org/2/library/datetime.html

[jreback]

related is #5519

[rockg]

It was added by #4706 and then subsequently removed to use the doc strings directly. @jorisvandenbossche, how does one document these methods/properties? Does one just need to add a weekday.doc?

In tseries.index

    dayofweek = _field_accessor('dayofweek', 'dow')
    weekday = dayofweek

[jorisvandenbossche]

@rockg Yes, I converted the manual list to a sphinx autosummary table, so it would also use the docstring. However, I did not get to the addition of the docstring. I will do now.

[jorisvandenbossche]

As a summary regarding this issue:

• The timestamp fields are coming from the python datetime module (which are documented there, see link that @jreback provided).
• The DatetimeIndex fields are now documented in api.rst with DOC: add way to document DatetimeIndex field attributes #5814 (at least for month and weekday)
• What could be done to close this issue:
  • Add Timestamp to api.rst with some explanation there about the fields (but problem with this, see DOC: documenting cython class (eg Timestamp): "cyfunction is not a python function" #5218).
  • Add a section to http://pandas.pydata.org/pandas-docs/dev/timeseries.html to explain Timestamp a little bit more (and that is is a subclass of datetime, so you can use all fields from datetime)

[jreback]

I think is for sure worthwhile to explain Timestamp a bit for in timeseries.rst (and maybe a mention in basics/10min, but meat in Timeseries)

(doesn't preclude your other buckets though )

[ghost]

related #3324

[jreback]

#6380 closed this