上个月的日期时间熊猫

Question

I have a datetime instance declared as follows:

dtDate = datetime.datetime(2016,1,1,0,0)

How do I get the previous month and previous year from dtDate ?

eg something like:

dtDate.minusOneMonth()
# to return datetime.datetime(2015,12,1,0,0)

Answer 1

You can use:

dtDate = datetime.datetime(2016,1,1,0,0)

print (dtDate - pd.DateOffset(months=1))
2015-12-01 00:00:00

print (dtDate - pd.DateOffset(years=1))
2015-01-01 00:00:00

Add s is important, because if use year only:

print (dtDate - pd.DateOffset(year=1))
0001-01-01 00:00:00 

Answer 2

You can use DateOffset :

In [32]:    
dtDate = dt.datetime(2016,1,1,0,0)
dtDate - pd.DateOffset(months=1)

Out[32]:
Timestamp('2015-12-01 00:00:00')

Answer 3

To Manipulate an entire pandas series.

Use pd.DateOffset() with .dt.to_period("M")

df['year_month'] = df['timestamp'].dt.to_period("M")
df['prev_year_month'] = (df['timestamp'] - pd.DateOffset(months=1)).dt.to_period("M")

If you want to go forward a month, set months=-1 .

Answer 4

Use relativedelta from dateutil :

import datetime
import dateutil.relativedelta
dtDate = datetime.datetime(2016,1,1,0,0)
# get previous month
print ((dtDate+dateutil.relativedelta.relativedelta(months=-1)).month)
# get previous year
print ((dtDate+dateutil.relativedelta.relativedelta(years=-1)).year)

Output:

12
2015