Analyze some Data

The dataset has 6 columns and around 3000 rows. Dates listed in sequence (for around three months). The values for the other 3 columns has whether 0 or 1 values. The program must do: 1- Divide each month to weeks, for example, the month February dates of days 1-7 replaced with week1, dates of days replaced for 8-14 week 2 etc.… 2- Sum the total of cases for each week for the dates in the dataset (three months). 3- The program only sum the first 1 in the week of a user and ignore the rest of other 1 in a week, for example: if the (user id: 48) in week 1(which means days between 01/02/2011-07/02/2011) has his first 1 in case1 then, we only count this one because it happens first then ignore the rest of 1 on that week. There is no duplication on the dataset. Any user can report any case out of these 4 cases, that’s why some users in same date reported different cases.