Pandas timestamp on array

Question

Pandas does not convert my array into an array of Timestamps:

a = np.array([1457392827660434006, 1457392828660434012,  1457392829660434023,1457474706167386148])
pd.Timestamp(a)

gives an an error :

TypeError                                 Traceback (most recent call last)
<ipython-input-42-cdf0e494942d> in <module>()
      1 a = np.array([1457392827660434006, 1457392828660434012, 1457392829660434023,1457474706167386148])
----> 2 pd.Timestamp(a)

pandas/tslib.pyx in pandas.tslib.Timestamp.__new__ (pandas/tslib.c:8967)()

pandas/tslib.pyx in pandas.tslib.convert_to_tsobject (pandas/tslib.c:23508)()

TypeError: Cannot convert input to Timestamp

Whereas looping on the array elements works just fine :

for i in range(4):
    t = pd.Timestamp(a[i])
    print t

gives :

2016-03-07 23:20:27.660434006
2016-03-07 23:20:28.660434012
2016-03-07 23:20:29.660434023
2016-03-08 22:05:06.167386148

As expected.

Moreover when that array is my first column in a csv file, it does not get parsed to a TimeStamp automatically, even if I specify parse_date correctly.

Any help please?

try: df = pd.to_datetime(a)

MaxU - stand with Ukraine
– MaxU - stand with Ukraine

2016-03-14 13:13:46 +00:00
Commented Mar 14, 2016 at 13:13 — MaxU - stand with Ukraine
– MaxU - stand with Ukraine, Commented Mar 14, 2016 at 13:13

jezrael · Accepted Answer · 2016-03-14 13:25:25Z

3

I think you can use to_datetime and then if you need array values:

import pandas as pd
import numpy as np

a = np.array([1457392827660434006, 1457392828660434012,  
              1457392829660434023,1457474706167386148])

print pd.to_datetime(a).values

['2016-03-08T00:20:27.660434006+0100' '2016-03-08T00:20:28.660434012+0100'
 '2016-03-08T00:20:29.660434023+0100' '2016-03-08T23:05:06.167386148+0100']

print pd.to_datetime(a, unit='ns').values
['2016-03-08T00:20:27.660434006+0100' '2016-03-08T00:20:28.660434012+0100'
 '2016-03-08T00:20:29.660434023+0100' '2016-03-08T23:05:06.167386148+0100']

edited Mar 14, 2016 at 13:25

answered Mar 14, 2016 at 13:13

jezrael

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

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

2 Comments

volatile Over a year ago

Thanks. It works but still does not explain the weird difference between vectorial call to pd.Timestamp versus calling it on each individual item.. And moreover does not explain why the column in the csv file does not get parsed automatically to a Timestamp, when I specify parse_dates

jezrael Over a year ago

I think pd.Timestamp not convert array to Timestamp, if unit are ns. For converting you can use function to_datetime.

Collectives™ on Stack Overflow

Pandas timestamp on array

1 Answer 1

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related