Editing a Pandas dataframe rows

Question

I have a large dataframe that I am trying to edit. I am trying to simplify the timestamp column by removing the date and the seconds so the column only shows a four-digit time e.g. "00:00". So far all I know what to do is use a for loop but I have no clue what condition to apply to this problem

              timestamp  date    activity   Id     total hour activity
720 2003-05-08 00:00:00 2003-05-08  0   condition_1 NaN
721 2003-05-08 00:01:00 2003-05-08  0   condition_1 NaN
722 2003-05-08 00:02:00 2003-05-08  0   condition_1 NaN
723 2003-05-08 00:03:00 2003-05-08  0   condition_1 NaN
724 2003-05-08 00:04:00 2003-05-08  0   condition_1 NaN
... ... ... ... ... ...
10794   2003-05-14 23:54:00 2003-05-14  0   condition_1 NaN
10795   2003-05-14 23:55:00 2003-05-14  12  condition_1 NaN
10796   2003-05-14 23:56:00 2003-05-14  0   condition_1 NaN
10797   2003-05-14 23:57:00 2003-05-14  18  condition_1 NaN
10798   2003-05-14 23:58:00 2003-05-14  0   condition_1 NaN
10079 rows × 5 columns

for i, row in df.iterrows():
if row['timestamp']:
    #delete row

What condition do you want to apply? Are you sure you need a condition, won't you be changing the value in the timestamp column of each row? — norie
– norie, Commented Mar 21, 2021 at 17:09

Anurag Dabas · Accepted Answer · 2021-03-21 17:42:42Z

2

Here I am assuming that the values of 'timestamp' column is of type string(Object)

You can do this by split() method:-

df['timestamp']=df['timestamp'].str.split(' ',expand=True)[1].str.split(':',1,expand=True)[1]

Explaination:- Firstly we splitted values by ' ' so

2003-05-08 00:00:00 is changed into [2003-05-08,00:00:00] and then we are selecting

'00:00:00' and again using split() method on it and breaking it by ':' so it becomes [00,00:00] and again we are grabbing 00:00 part from it

And If the 'timestamp' column is of datetime dtype then you can use:-

df['timestamp']=df['timestamp'].dt.strftime('%H:%M')

edited Mar 21, 2021 at 17:42

answered Mar 21, 2021 at 17:13

Anurag Dabas

24.3k9 gold badges25 silver badges41 bronze badges

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

Comments

sofktrn · Accepted Answer · 2021-03-21 17:29:22Z

0

In order to keep hour and minutes, you could do the following:

df['timestamp'] = pd.to_datetime(df.timestamp)  # this step only if the "timestamp" column is not already datetime

df['timestamp']=df['timestamp'].dt.strftime('%H:%M')  # keeps hours (H) and minutes(M)

answered Mar 21, 2021 at 17:29

sofktrn

563 bronze badges

Comments

Ostap Orishko · Accepted Answer · 2021-03-21 17:17:53Z

0

You don't need a for loop to grab hours and minutes only. If your column is datetime, then you can create a separate column from your timestamp by calling:

df['Time'] = df.timestamp.apply(lambda x: str(x.time())[:-3])

That will return "00:00" format as a string. Alternatively you can just call

df['Time'] = df.timestamp.apply(lambda x: x.time())

To get "00:00:00" format of type datetime.time.

answered Mar 21, 2021 at 17:17

Ostap Orishko

711 silver badge6 bronze badges

Comments

norie · Accepted Answer · 2021-03-21 17:26:15Z

0

If you want a column, say 'Time', with only the time from the timestamp you can use this.

df['Time'] = df['timestamp'].dt.time

answered Mar 21, 2021 at 17:26

norie

9,9372 gold badges14 silver badges19 bronze badges

Collectives™ on Stack Overflow

Editing a Pandas dataframe rows

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related