Convert a timestamp to int using astype(int)

Question

I have the variable time:

>>> time = pd.to_datetime(1613260800000000000)
>>> time
Timestamp('2021-02-14 00:00:00')

time is a timestamp. Now I want to convert time back to an int value. But Timestamp object has no attribute astype. Running the following code:

>>>time.astype(int)

AttributeError: 'Timestamp' object has no attribute 'astype'

I know that for a column of a dataframe I can do:

>>> df['start_time'] = df['start_time'].to_timedate()
>>> df['start_time'] = df['start_time'].astype(int)

The second command assigns the type of the column to int.

But I did not find a comprehensive source explaining what to do with individual Timestamp argument.

How can I solve this issue?

Hi, may I know what is the expected output? – mpx Mar 05 '21 at 06:12 — mpx, Mar 05 '21 at 06:12

score 4 · Answer 1 · answered Mar 05 '21 at 06:14

4

just use time_variable.timestamp()

but that will be in seconds ... instead of ms or whatever (us?)

you will need to multiply the result by 1e9 to get back the same value as you put in

answered Mar 05 '21 at 06:14

Joran Beasley

110,522
12
160
179

1

`time_variable.value` returns the timestamp as nanoseconds since the epoch (as the OPs input). – FObersteiner Mar 05 '21 at 07:36

jezrael · Accepted Answer · 2021-03-05T06:20:21.253

For ns like native format for numpy/pandas use:

d = pd.to_datetime(1613260800000000000)

native = int(d.timestamp() * 10**9)
print (native)
1613260800000000000

print (pd.to_datetime(native))
2021-02-14 00:00:00

If need convert column:

time = pd.to_datetime(1613260800000000000)

df = pd.DataFrame({'start_time':[time, time]})

print (df['start_time'].astype(np.int64))
0    1613260800000000000
1    1613260800000000000
Name: start_time, dtype: int64

Convert a timestamp to int using astype(int)

2 Answers2