CODEWITHSUNDEEP
pythonpandas
Morello
Morello

Reputation: 258

pandas - datetime column subtraction datetime strptime

I have a column creation_date in my pandas dataframe as type str that has the following format:

2020-02-06 11:35:17+00:00

I am trying to create a new column in my dataframe titled days_since_creation

from datetime import date

today = date.today()

df['days_since_creation'] = date.strptime(df['creation_date'],'%m/%d/%Y') - today

My code is not correct and i'm not sure where I am going wrong. any help is much appreciated

Upvotes: 1

Views: 632

Answers (2)

Suraj Shejal
Suraj Shejal

Reputation: 650

this will work

df['days_since_creation'] = pd.to_datetime(df['creation_date'])
df['days_since_creation'] = today - df.days_since_creation.dt.date

if you want just integer of no of days

df['days_since_creation'] = (today - df.days_since_creation.dt.date ).dt.days

Upvotes: 1

mechanical_meat
mechanical_meat

Reputation: 169324

Just tested this out and works alright for me, apologies for the error from my comment.

df['days_since_creation'] = (pd.to_datetime(df['creation_date']).dt.date - 
                             pd.Timestamp.today().date())

Result w just that one sample:

               creation_date days_since_creation
0  2020-02-06 11:35:17+00:00           -646 days

Edit: if the granularity is important and you want to specify how to handle the timezone information, do something like this instead:

df['days_since_creation'] = (pd.to_datetime(df['creation_date']).dt.tz_localize(None) - 
                             pd.Timestamp.today())

Result of that:

               creation_date        days_since_creation
0  2020-02-06 11:35:17+00:00 -647 days +23:26:51.569694

Upvotes: 1

Related Questions