Python: transform float into datetime timestamp

Question

How could I transform the following float:

9.3125

which corresponds to a time, into a proper Pandas timestamp in the likes of:

Timestamp('2017-11-13 10:00:00')

The float belongs to the same day, month, and year as the example timestamp above. I have been trying this:

from datetime import datetime
datetime.fromtimestamp(9.3125).strftime('%Y-%m-%d %H')

but this returns:

OSError: [Errno 22] Invalid argument

jezrael · Accepted Answer · 2018-09-12 11:52:59Z

2

num = 9.3125
a = pd.Timestamp('2017-11-13')  + pd.to_timedelta(num, unit='H').ceil('H')
print (a)
2017-11-13 10:00:00

answered Sep 12, 2018 at 11:46

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

But 9.3125 corresponds to about 09:19, so this solution would let me lose those 19 minutes

If I use float(num) instead of int(num) I get: ValueError: time data '2017-11-13 9.3125' doesn't match format specified.

FaCoffee · Accepted Answer · 2018-09-12 11:56:16Z

1

Inspired by the answers above, I have verified that the following works for me:

num = 9.3125
a = pd.Timestamp('2017-11-13')  + pd.to_timedelta(num, unit='H').ceil('min')
print (a)
Timestamp('2017-11-13 09:19:00')

answered Sep 12, 2018 at 11:56

FaCoffee

7,97933 gold badges105 silver badges184 bronze badges