foxtrot
01-16-2009, 07:52 AM
I have four variables: S_No, a, b, c for an n = 100.
Each new S_No is a new, unique case.
There are no repeat measurements per S_No.
a is the variable of interest and I sort the entire data-set "by a" (default ascending).
Now what I want to build, given data is sorted "by a" is:
d = a2 - a1 (S_no = 1).... a100-a99 (S_No = 100)
(d1 can = a1, but d2...d99 should be a2-a1...a100-a99)
Likewise,
e = b2 - b1 (S_no = 1).... b100-b99 (S_No = 100)
f = c2 - c1 (S_no = 1).... c100-c99 (S_No = 100)
Thus, total n will decrease my n-1.
I get three delta values,
Data sorted by a.
How to do this? :shakehead
Many thanks whoever gets me sorted out. :)
Each new S_No is a new, unique case.
There are no repeat measurements per S_No.
a is the variable of interest and I sort the entire data-set "by a" (default ascending).
Now what I want to build, given data is sorted "by a" is:
d = a2 - a1 (S_no = 1).... a100-a99 (S_No = 100)
(d1 can = a1, but d2...d99 should be a2-a1...a100-a99)
Likewise,
e = b2 - b1 (S_no = 1).... b100-b99 (S_No = 100)
f = c2 - c1 (S_no = 1).... c100-c99 (S_No = 100)
Thus, total n will decrease my n-1.
I get three delta values,
Data sorted by a.
How to do this? :shakehead
Many thanks whoever gets me sorted out. :)