Return the [Date] that within the last 14 days

[bnadir55]

Hi, I have a column [Date] in df that return date in this format : 20220628T065216.437 GMT
I need to return the [Date] that within the last 3 days,
tried with existing format : 20220628T065216.437 GMT
date = [Date] >= datetime.now() - timedelta(hours=72, minutes=0)]
but received errors , can I still use it somehow or do I need to convert it , is so can you please help with how ? Thx!

[deanhystad]

Please show what you have tried.

[bnadir55]

Please show what you have tried.

date = [Date] >= datetime.now() - timedelta(hours=72, minutes=0)]

[snippsat]

Can do it like this and Pandas has it own Date/Time stuff build in.

import pandas as pd
from io import StringIO

data = StringIO('''\
Date Tval Rnum Min Blend Result
12/7/2022 0.000091 2.5.2 9 0.1 25.0
12/6/2022 0.000131 1.9.7 5 0.5 50.0
12/5/2022 0.000131 1.2.44 1.2 150.0
12/4/2022 0.000555 1.3.44 1.4 17.0''')

df = pd.read_csv(data, sep=' ')
# Make sure that it is Pandas <datetime64[ns]> format
df['Date']= pd.to_datetime(df['Date'])

# From today
>>> df[df["Date"] > (pd.to_datetime('today') - pd.Timedelta(days=3))]
        Date      Tval    Rnum  Min  Blend  Result
0 2022-12-07  0.000091   2.5.2  9.0    0.1    25.0
1 2022-12-06  0.000131   1.9.7  5.0    0.5    50.0
2 2022-12-05  0.000131  1.2.44  1.2  150.0     NaN

# Use fixed date
>>> df[df["Date"] > (pd.to_datetime('2022-12-7') - pd.Timedelta(days=3))]
        Date      Tval    Rnum  Min  Blend  Result
0 2022-12-07  0.000091   2.5.2  9.0    0.1    25.0
1 2022-12-06  0.000131   1.9.7  5.0    0.5    50.0
2 2022-12-05  0.000131  1.2.44  1.2  150.0     NaN

[bnadir55]

Can do it like this and Pandas has it own Date/Time stuff build in.

import pandas as pd
from io import StringIO

data = StringIO('''\
Date Tval Rnum Min Blend Result
12/7/2022 0.000091 2.5.2 9 0.1 25.0
12/6/2022 0.000131 1.9.7 5 0.5 50.0
12/5/2022 0.000131 1.2.44 1.2 150.0
12/4/2022 0.000555 1.3.44 1.4 17.0''')

df = pd.read_csv(data, sep=' ')
# Make sure that it is Pandas <datetime64[ns]> format
df['Date']= pd.to_datetime(df['Date'])

# From today
>>> df[df["Date"] > (pd.to_datetime('today') - pd.Timedelta(days=3))]
        Date      Tval    Rnum  Min  Blend  Result
0 2022-12-07  0.000091   2.5.2  9.0    0.1    25.0
1 2022-12-06  0.000131   1.9.7  5.0    0.5    50.0
2 2022-12-05  0.000131  1.2.44  1.2  150.0     NaN

# Use fixed date
>>> df[df["Date"] > (pd.to_datetime('2022-12-7') - pd.Timedelta(days=3))]
        Date      Tval    Rnum  Min  Blend  Result
0 2022-12-07  0.000091   2.5.2  9.0    0.1    25.0
1 2022-12-06  0.000131   1.9.7  5.0    0.5    50.0
2 2022-12-05  0.000131  1.2.44  1.2  150.0     NaN

Thx!
however, not sure I understand your solution ,let me clarify,
I have this column value df[Date] ,and I tried to use it in here :

Date = df[Date]
df = Date > (pd.to_datetime('today') - pd.Timedelta(days=3))

and I am greeting error :
TypeError: '>' not supported between instances of 'str' and 'Timestamp'


can you please tell me what to replace with what ?
Thx!

[snippsat]

I have this column value df[Date] ,and I tried to use it in here :

Date = df[Date]

You can not do this,and that's not what i am doing.
Like this:

df['Date'] = pd.to_datetime(df['Date'])

From start and always check dtypes.

import pandas as pd
from io import StringIO

data = StringIO('''\
[python]Date Tval Rnum Min Blend Result
12/7/2022 0.000091 2.5.2 9 0.1 25.0
12/6/2022 0.000131 1.9.7 5 0.5 50.0
12/5/2022 0.000131 1.2.44 1.2 150.0
12/4/2022 0.000555 1.3.44 1.4 17.0''')

df = pd.read_csv(data, sep=' ')

Use.

>>> df
        Date      Tval    Rnum  Min  Blend  Result
0  12/7/2022  0.000091   2.5.2  9.0    0.1    25.0
1  12/6/2022  0.000131   1.9.7  5.0    0.5    50.0
2  12/5/2022  0.000131  1.2.44  1.2  150.0     NaN
3  12/4/2022  0.000555  1.3.44  1.4   17.0     NaN

>>> df.dtypes
Date       object
Tval      float64
Rnum       object
Min       float64
Blend     float64
Result    float64
dtype: object

See that Date is object,this is not correct should be datetime.

>>> df['Date'] = pd.to_datetime(df['Date'])
>>> df.dtypes
Date      datetime64[ns]
Tval             float64
Rnum              object
Min              float64
Blend            float64
Result           float64
dtype: object

No is Date column ok,and then it will work.
As one day has past get 2.

>>> df[df["Date"] > (pd.to_datetime('today') - pd.Timedelta(days=3))]
        Date      Tval   Rnum  Min  Blend  Result
0 2022-12-07  0.000091  2.5.2  9.0    0.1    25.0
1 2022-12-06  0.000131  1.9.7  5.0    0.5    50.0

>>> df[df["Date"] > (pd.to_datetime('today') - pd.Timedelta(days=4))]
        Date      Tval    Rnum  Min  Blend  Result
0 2022-12-07  0.000091   2.5.2  9.0    0.1    25.0
1 2022-12-06  0.000131   1.9.7  5.0    0.5    50.0
2 2022-12-05  0.000131  1.2.44  1.2  150.0     NaN