How to create new column with conditional datetime using pandas

Question

I am trying to add a new column containing label with this condition:

Label 1 if delta time between value in 'time' and dt < 2 hours
Label 0 for other case

My current idea:

df = pd.read_csv('./datetimecek.csv')
df['time'] = pd.to_datetime(df['datetime'])

dt = datetime.strptime("19/02/18 19:00", "%d/%m/%y %H:%M")

datetime            time
2018/02/19 16:00    2018-02-19 16:00:00
2018/02/19 17:00    2018-02-19 17:00:00
2018/02/19 18:00    2018-02-19 18:00:00
2018/02/19 19:00    2018-02-19 19:00:00

And then I defined timedelta

a = timedelta(hours=2)

def label(c):
if dt - df['time'] < a:
    return '1'
else:
    return '0'

then

df['label'] = df.apply(label, axis=1)

But I got error: 'The truth value of a Series is ambiguous. Use a.empty, a.bool()...

Is there anyway I can fix this?

I think you meant to use c in the function definition of label instead of the entire df existing in global scope. — crazyGamer
– crazyGamer, Commented Jan 29, 2019 at 6:15

jezrael · Accepted Answer · 2019-01-29 06:24:00Z

1

If want set strings 0 and 1:

df['label'] = np.where(dt - df['time'] < a, '1','0')

Or alternative by @Dark:

df['label'] = (dt - df['time'] < a).astype(int).astype(str)

print (df)
           datetime                time label
0  2018/02/19 16:00 2018-02-19 16:00:00     0
1  2018/02/19 17:00 2018-02-19 17:00:00     0
2  2018/02/19 18:00 2018-02-19 18:00:00     1
3  2018/02/19 19:00 2018-02-19 19:00:00     1

print (type(df.loc[0, 'label']))
<class 'str'>

If want set integers 0 and 1:

df['label'] = (dt - df['time'] < a).astype(int)

Alternative:

df['label'] = np.where(dt - df['time'] < a, 1,0)

print (df)
           datetime                time label
0  2018/02/19 16:00 2018-02-19 16:00:00     0
1  2018/02/19 17:00 2018-02-19 17:00:00     0
2  2018/02/19 18:00 2018-02-19 18:00:00     1
3  2018/02/19 19:00 2018-02-19 19:00:00     1

print (type(df.loc[0, 'label']))
<class 'numpy.int32'>

Is there anyway I can fix this?

Yes, need change df to c for working with scalars:

def label(c):
    if dt - c['time'] < a:
        return '1'
    else:
        return '0'

edited Jan 29, 2019 at 6:24

answered Jan 29, 2019 at 6:12

jezrael

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

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

1 Comment

Bharath M Shetty Over a year ago

I was close np.where(dt-df['time']<a).astype(int).astype(str) :)

Collectives™ on Stack Overflow

How to create new column with conditional datetime using pandas

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related