The dataframe contains date column, revenue column(for specific date) and the name of the day.

This is the code for creating the df:

pd.DataFrame({'Date':['2015-01-08','2015-01-09','2015-01-10','2015-02-10','2015-08-09','2015-08-13','2015-11-09','2015-11-15'],
             'Revenue':[15,4,15,13,16,20,12,9],
             'Weekday':['Monday','Tuesday','Wednesday','Monday','Friday','Saturday','Monday','Sunday']})

I want to find the sum of revenue between Mondays:

2015-02-10   34  Monday
2015-11-09   49  Monday  etc.

CodePudding user response：

First idea is used Weekday for groups by compare by Monday with cumulative sum and aggregate per groups:

df1 = (df.groupby(df['Weekday'].eq('Monday').cumsum())
        .agg({'Date':'first','Revenue':'sum', 'Weekday':'first'}))
print (df1)
              Date  Revenue Weekday
Weekday                            
1       2015-01-08       34  Monday
2       2015-02-10       49  Monday
3       2015-11-09       21  Monday

But seems not matched Weekday column with Dates in sample data, so DataFrame.resample per weeks starting by Mondays return different output:

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

df2 = df.resample('W-Mon', on='Date').agg({'Revenue':'sum', 'Weekday':'first'}).dropna()
    
print (df2)
            Revenue   Weekday
Date                         
2015-01-12       34    Monday
2015-02-16       13    Monday
2015-08-10       16    Friday
2015-08-17       20  Saturday
2015-11-09       12    Monday
2015-11-16        9    Sunday

CodePudding user response：

First convert your Date column from string to datetime type:

df.Date = pd.to_datetime(df.Date)

Then generate the result:

result = df.groupby(pd.Grouper(key='Date', freq='W-MON', label='left')).Revenue.sum()/
    .reset_index()

This result does not contain day of week and in my opinion this is OK, as they will be all Mondays.

If you want to see only weeks with non-zero result, you can get it as:

result[result.Revenue != 0]

For your source data the result is:

         Date  Revenue
0  2015-01-05       34
5  2015-02-09       13
30 2015-08-03       16
31 2015-08-10       20
43 2015-11-02       12
44 2015-11-09        9

CodePudding user response：

happy with what

feelings heart

you have

1. no regrets
2. no crying
3. smiling face