CODEWITHSUNDEEP
pythondatetimepandas
Tasos
Tasos

Reputation: 7577

Find how many weeks passed in Python Pandas

I have a Pandas dataframe with a datetime column. My problem is the following:

I have a starting date of 04/08/2014. Since then, I count weeks in chunks of 16 weeks. So, from 04/08/2014 until 11/08/2014, it will be week 1. After 16 weeks, it will start again from week 1. I want to create a new column where it will find the week of the current chunk based on the datetime column.

This is what I have done, but it seems that it doesn't work as it should.

startingweek = datetime.date(2014, 8, 4)

df['WeekChunk'] = int(((df['DateTimeColumn'] - startingweek) / pd.offsets.Day(1))/7/16)

I calculated the number of days between the two days, then divided by 7 days to find the number of weeks and then divided by 16 to find the week of chunk.

If I use a date of 23/12/2015, it should be week 9. But, the above code seems wrong.

Upvotes: 0

Views: 1091

Answers (2)

Steve Misuta
Steve Misuta

Reputation: 1033

Here is a way to do it using built-in numpy/pandas time series functionality without using modulo operator:

import pandas as pd
import numpy as np

# re-create a dummy df with a Date column for this example
startingweek = datetime.date(2014, 8, 4)
df = pd.DataFrame(pd.date_range(startingweek, periods=1000, freq='D'), columns=['Date'])

# calc the WeekChunks by recasting the differences into timedelta periods
df['WeekChunks'] = 1 + (df.Date-startingweek).astype(np.timedelta64(1,'W'))
                     - (df.Date-startingweek).astype(np.timedelta64(1,'16W'))*16

# find Dec 23, 2015...should be WeekChunks = 9
df.set_index('Date').ix['2015-12-23']

WeekChunks    9
Name: 2015-12-23 00:00:00, dtype: float64

Upvotes: 0

Leonid Mednikov
Leonid Mednikov

Reputation: 973

If you need the week in a period of 16, you need the modulo, not devision. So change "/" to "%". And get int() before that.

df['WeekChunk'] = int(((df['DateTimeColumn'] - startingweek) / pd.offsets.Day(1))/7) % 16

P.S. But the first week would be 0, not 1.

Upvotes: 1

Related Questions