More summarizing....

[mayday1]

OK, I've got a long list of patients. Each row show ID Number, start date, end date (which may be blank if treatment is still going on). An ID number may appear once or in 20 rows for reasons that aren't relevant here. If an ID number does appear more than once, the start date and end dates area always the same and everything else can be ignored.

Sample data:
111111111, 1/1/14, 1/5/14
111111111, 1/1/14, 1/5/14
111111111, 1/1/14, 1/5/14
222222222, 1/1/14,
222222222, 1/1/14,
333333333, 1/3/14, 1/8/14
333333333, 1/3/14, 1/8/14
444444444, 1/2/14. 1/12/14
555555555, 1/5/14.

What I need is a summary that shows, for every date starting with the earliest start date in the file and going up to the current date:
The Date and:
- How many of these records were started on this date?
- How many were flagged as closed on this date?
- What is the highest number of records open at any time on this date?
- What is the average number of days that those open at any point on this date existed?
- What is the median number of days that those open at one point on this date existed? This one seems especially hard and I could do this manually if nobody can figure it out.
- How old is the oldest record that was open at some point on this date?
- What is the lowest number in days that a record was open on this date?

So for the Sample data above it would look something like this: (I’m making up all the numbers)
1/1/14, 5, 0, 5, 1, 1, 1,1
1/2/14 1,0,6,.8,2,2.1
1/3/14 2,0,8,2.4,3,3,1
1/4/14 0,0,8,3, 4,4,2 (There were no records started on this date but I still need the data for all other records.)
1/5/14 etc....

 

[mayday1]

Possibly, if you explain things as requested in post #4, for example...

Yes - I was using the examples I posted and doing the math in my (air) head.

 

[Aladin Akyurek]

Yes - I was using the examples I posted and doing the math in my (air) head.

What I'm asking for is how those results obtain given your sample, i.e., the manual procedures/calculations that you have used.