convert yearweek format to date python

Question

i have a dataframe like this (this data in bigquery. And yearweek format is %Y%U (weeks starts sunday):

data = {'yearweek': ['202140', '202139', '202138', '202137'], 'value': [452, 741, 475, 1000]}
df=pd.DataFrame(data)

yearweek   value
202140     452
202139     741
202138     475
202137     1000

expected output:

yearweek   value  date
202140     452    04-10-2021
202139     741    27-09-2021
202138     475    20-09-2021
202137     1000   13-09-2021

i tried this:

df['date'] = pd.to_datetime(df.yearweek + '0', format='%Y-%W%w')

but getting this error:

ValueError: time data '2021420' does not match format '%Y-%W%w' (match)

tried too:

df['datex'] = pd.to_datetime(df.yearweek , format='%Y-%W%w')

but error again

ValueError: time data '202142' does not match format '%Y-%W%w' (match)

Answer 1

You can use format as format='%Y%U%w'

>>> import pandas as pd
>>> df = pd.DataFrame({'yearweek': ['202140', '202139', '202138', '202137'], 'value': [452, 741, 475, 1000]})
>>> df
  yearweek  value
0   202140    452
1   202139    741
2   202138    475
3   202137   1000
>>> df['date'] = pd.to_datetime(df.yearweek + '0', format='%Y%U%w')
>>> df
  yearweek  value       date
0   202140    452 2021-10-03
1   202139    741 2021-09-26
2   202138    475 2021-09-19
3   202137   1000 2021-09-12