pandas datetime时间处理一学就会

Overview

Definition	Element type	Series type	Pandas data type
Date times	Timestamp	DatetimeIndex	datetime64[ns]
Time deltas	Timedelta	TimedeltaIndex	timedelta64[ns]
Time spans	Period	PeriodIndex	period[freq]
Date offsets	DateOffset	None	None

Date times

Timestamp

pd.Timestamp(year, month, day, hour, minute, second)
pd.Timestamp(ts_input)

Keyword arguments year, month, day, hour, minute, second,
used to determine a timestamp.
ts_input is a time str.

Instance attributes

Attributes year, month, day, hour, minute, second,
get the time.
value, the num of nanoseconds
from zero on January 1, 1970(Computer Year)
to the given timestamp.

Class attributes

max, min, the max and min timestamp.

Datetime series

to_datetime

pd.to_datetime(arg, formatstr, )

arg, int, float, str, datetime, list, tuple, 1-d array, Series, DataFrame/dict-like
The object to convert to a datetime.
If a DataFrame is provided, the method expects minimally the following columns: “year”, “month”, “day”.

errors, {‘ignore’, ‘raise’, ‘coerce’}, default ‘raise’.
For a invalid parsing,

• ‘raise’, raise an exception.
• ‘coerce’, set as NaT.
• ‘ignore’, return the input.

formatstr, default None. The strftime to parse time.
Refer to strftime documentation.

Range

pd.date_range(start=None, end=None, 
              periods=None, freq=None,
              tz=None, normalize=False,
              name=None, inclusive='both',
              *, 
              unit=None, **kwargs)

freq is the frequence, auto specified as ‘D’.
tz is time zone.
normalize, bool, default False,
whether normalize start/end dates to midnight.
inclusive, {‘both’, ‘neither’, ‘left’, ‘right’}, default ‘both’,
whether to set each bound as closed or open.

bdate_range, used to gen a series without holidays.

pd.bdate_range(start=None, end=None,
               periods=None, freq='B',
               tz=None, normalize=True,
               name=None,
               weekmask=None, holidays=None,
               inclusive='both',
               **kwargs)

freq, must specified.
weekmask, str,
mask of valid business days,
passed to numpy.busdaycalendar
auto specified as ‘Mon Tue Wed Thu Fri’.
holidays, list-like,
dates to exclude from the set of valid business days,
passed to numpy.busdaycalendar.

Change freq

s.asfreq(freq, 
         method=None, how=None,
         normalize=False,
         fill_value=None)

method, {‘backfill’/‘bfill’, ‘pad’/‘ffill’}, default None
how, {‘start’, ‘end’}, default end, for PeriodIndex only.
fill_value, scalar,
value to use for missing values, applied during upsampling.

dt object

s.dt

Normal attributes,
date, time, year, month, day, hour, minute, second, microsecond, nanosecond, dayofweek, dayofyear, weekofyear, daysinmonth, quarter

Returning a bool value,

s.dt.is_year_start
s.dt.is_quarter_start
s.dt.month_start

s.dt.is_year_end
s.dt.is_quarter_end
s.dt.is_month_end

Round,

s.dt.round(freq)
s.dt.ceil(freq)
s.dt.floor(freq)

Timedelta

Generate

Generate a pd.Timedelta,

pd.Timedelta(days, months, ...)
pd.Timedelta(time_str)

timestamp1 - timestamp2 # return a pd.Timedelta

Generate a pd.Series,

pd.to_timedelta(s)
pd.timedelta_range(start, end, freq, period)

Calculation

timestamp1 - timestamp2 # return a pd.Timedelta
timestamp + timedelta # return a pd.Timestamp

Offset

Offset methods

We could check all offset methods, using

pd.offsets.__all__

There are some usual offset methods.

pd.offsets.Day()
pd.offsets.BDay()
pd.offsets.CDay()
pd.offsets.WeekofMonth()
pd.offsets.MonthBegin()
pd.offsets.MonthEnd()

Offset strings

We could use offset strings to denote offsets as parameter freq.

String	Offset
‘MS’	MonthBegin()
'M"	MonthEnd()
‘B’	BDay()
‘W-MON’	CDay(weekmask=‘Mon’)
‘WOM-1MON’	WeekOfMonth(week=0,weekday=0)

More refer to Offset aliases.

Date rolling and resample

Rolling

When the data type of index is datetime64[ns], we could use date rolling.

s.rolling(time)

It has same attributes and methods with normal rolling.

Shift

A slide method of series with time index.

s.shift(freq)

Resample

It is same with groupby, specified to datetime series.

s.resample(rule, origin)

rule, DateOffset, Timedelta or str,
the offset string or object representing target conversion.
origin, {‘epoch’, ‘start’, ‘start_day’},
Timestamp or str, default ‘start_day’.

• ‘epoch’: 1970-01-01
• ‘start’: the first value of the timeseries
• ‘start_day’: the first day at midnight of the timeseries