Month column in dataframe, plus number of months, to compare with current calendar month

Question

A column in dataframe looks like month, I want to use it to plus a month, as a ‘future’ month, then to compare this ‘future’ month with current (calendar) month.

import pandas as pd
from io import StringIO
import numpy as np
from datetime import datetime

csvfile = StringIO(
"""Name Year - Month    Score
Mike    2022-11 31
Mike    2022-09 136
""")
df = pd.read_csv(csvfile, sep = 't', engine='python')

d_name_plus_month = {"Mike":2}

month_of_first_row = pd.to_datetime(df.iloc[[0]]['Year - Month']).values.astype("datetime64[M]")

plus_months = d_name_plus_month['Mike']

scheduled_month = month_of_first_row + int(plus_months)
# scheduled_month_in_string = scheduled_month.astype('str')

current_month = datetime.now().strftime("%Y") +'-' +datetime.now().strftime("%m")        # it's string
current_month = np.array(current_month)

print (scheduled_month <=  current_month)

# month_of_first_row: 2022-11
# scheduled_month: 2023-01
# current_month: 2023-02
# so "scheduled_month" is earlier than "current_month".

But it has error:

TypeError: '<=' not supported between instances of 'numpy.ndarray' and 'numpy.ndarray'

I’ve tried to alert the lines to make them into string for compare, but not successful.

How can I correct the lines?

Asked By: Mark K

||

Source

Answer 1

I suggest use month periods by Serie.dt.to_period for easy add/ remove months by integers:

d_name_plus_month = {"Mike":2}

month_of_first_row = pd.to_datetime(df['Year - Month'].iat[0]).to_period('m')
print (month_of_first_row)
2022-11

plus_months = d_name_plus_month['Mike']
scheduled_month = month_of_first_row + int(plus_months)

current_month = pd.Timestamp.now().to_period('m')
print (current_month)
2023-02

print (scheduled_month <=  current_month)
True

Answered By: jezrael

Answer 2

You can convert both scheduled_month and current_month to strings before comparison

scheduled_month_str = scheduled_month.astype('datetime64[M]').astype(str)

current_month_str = datetime.now().strftime("%Y-%m")

print(scheduled_month_str <= current_month_str)

Answered By: angwrk

Answer 3

month_of_first_row is an array with one value, you can add a [0] to convert it to get a single datetime object

month_of_first_row = pd.to_datetime(df.iloc[[0]]['Year - Month']).values.astype("datetime64[M]")[0]

And then convert the current month to a datetime objetc as well instead of a numpy array

# no: current_month = np.array(current_month)
current_month = pd.to_datetime(current_month) # yes

Answered By: muhmann

Answer 4

seems like right now
current month is np.array('2023-02', dtype='<U7')
and scheduled month is np.array(['2023-01'], dtype='datetime64[M]')

using this row will bring them to the same type:

current_month =np.array([current_month], dtype='datetime64[M]')

which should solve the error.

full code:

    import pandas as pd
    from io import StringIO
    import numpy as np
    from datetime import datetime
    
    csvfile = StringIO(
    """Name,Year - Month,Score
    Mike,2022-11,31
    Mike,2022-09,136
    """)
    df = pd.read_csv(csvfile, sep = ',', engine='python')
    
    d_name_plus_month = {"Mike":2}
    
    month_of_first_row = pd.to_datetime(df.iloc[[0]]['Year - Month']).values.astype("datetime64[M]")
    
    plus_months = d_name_plus_month['Mike']
    
    scheduled_month = month_of_first_row + int(plus_months)
    # scheduled_month_in_string = scheduled_month.astype('str')
    
    current_month = datetime.now().strftime("%Y") +'-' +datetime.now().strftime("%m")        # it's string
    current_month =np.array([current_month], dtype='datetime64[M]')

    print (scheduled_month <=  current_month)

Answered By: Manny

Month column in dataframe, plus number of months, to compare with current calendar month

Question:

Answers: