CODEWITHSUNDEEP
pythonpandastimedelta
Octavarium
Octavarium

Reputation: 131

Convert nanoseconds to days, hours, min and seconds

I'm working in python with a pandas df and trying to convert a column that contains nanoseconds to a time with days, hours, minutes and seconds, but I'm not succeeding.

My original df looks like this:

ID         TIME_NANOSECONDS     
1          47905245000000000                       
2          45018244000000000                       
3          40182582000000000

The result should look like this:

ID         TIME_NANOSECONDS          TIME
1          47905245000000000         554 days 11:00:45.000000000               
2          45018244000000000         521 days 01:04:04.000000000 
3          40182582000000000         465 days 01:49:42.000000000

I've found some answers that advised to use timedelta, but following code, returns a date, which is not what I want.

temp_fc_col['TIME_TO_REPAIR_D'] = datetime.timedelta(temp_fc_col['TIME_TO_REPAIR'], unit='ns')

alternatively,

temp_fc_col['TIME_TO_REPAIR_D'] = (timedelta(microseconds=round(temp_fc_col['TIME_TO_REPAIR'], -3)

Returns an error: unsupported type for timedelta microseconds component: Series. Probably, because this staement can only process one value at a time.

Upvotes: 3

Views: 4363

Answers (1)

jezrael
jezrael

Reputation: 863116

Use to_timedelta working well with Series, also unit='ns' should be omit:

temp_fc_col['TIME_TO_REPAIR_D'] = pd.to_timedelta(temp_fc_col['TIME_NANOSECONDS'])
print (temp_fc_col)
   ID   TIME_NANOSECONDS  TIME_TO_REPAIR_D
0   1  47905245000000000 554 days 11:00:45
1   2  45018244000000000 521 days 01:04:04
2   3  40182582000000000 465 days 01:49:42

Upvotes: 3

Related Questions