QST: DatetimeIndex changes

[mboldin-temple]

• I have searched the [pandas] tag on StackOverflow for similar questions.

• I have asked my usage related question on StackOverflow.

---

Question about pandas

In version 1.0.5, DatetimeIndex seems greatly changed from earlier versions
Was this intentional and documented in the version changes?

Many statsmodels forecasting steps now break

For example, the code below no longer works

dx = pd.DatetimeIndex(start ='2020-01-01 00:00', freq ='M', periods = 12)

I get
TypeError: new() got an unexpected keyword argument 'start'

The documentation has these arguments

class pandas.DatetimeIndex(data=None, freq=None, tz=None, normalize=False, closed=None, ambiguous='raise', dayfirst=False, yearfirst=False, dtype=None, copy=False, name=None)[source]

with little explanation or usefulness IMHO

I see old Pandas has
def new(cls, data=None,
freq=None, start=None, end=None, periods=None, tz=None,
normalize=False, closed=None, ambiguous='raise',
dayfirst=False, yearfirst=False, dtype=None,
copy=False, name=None, verify_integrity=True):

# Your code here, if applicable
# Create the DatetimeIndex 
dx = pd.DatetimeIndex(start ='2020-01-01 00:00', freq ='M',  periods = 12)
## expect TypeError: __new__() got an unexpected keyword argument 'start'

[mboldin-temple]

I see now it is documented in What's New
Removed the previously deprecated keywords “start”, “end”, and “periods” from the DatetimeIndex, TimedeltaIndex, and PeriodIndex constructors; use date_range(), timedelta_range(), and period_range() instead (GH23919)

and date_range() works as I need it

Still some statsmodels routines will fail when expecting to use start= as a keyword

[jreback]

these were deprecated in 0.24.0 and removed in 1.0.x. you should push statsmodels to update this.

[jreback]

xref #23919