Need suggestion in method to use

Question

Greeting.
I'm doing data mining stuff and the data look like this:
a     c
1     1
2     1
3     1
7     1
8     1
6     2
7     2
8     2
9     2
10     2This data will be the input. I need to calculate a 'split-point' in 'a' according to class 'c' i.e. 1 and 2. I wanna group 'a' into two class. Let say after calculation the split point is 8, then output will be {1,2,3,6,7,8} and {9,10} for 'a' means 1,2,3,6,7 and 8 will become {1,2,3,6,7,8}
First I need to sort 'a'. 'a' and 'c' are parallel. And then I need to calculate probability of 'c' for a point let say 8, number <=8 that is '1' in 'c'. In this case will be 5/8.
This is some description.
My problem is: Which way will be easier to do this? Using what kind of input? (excel, txt, csv) ?
How to get start?
Edited by: Raiko on Apr 26, 2010 9:18 PM
Edited by: Raiko on Apr 26, 2010 9:20 PM
Edited by: Raiko on Apr 26, 2010 9:23 PM

Answer 1

Raiko wrote:
This data will be the input. I need to calculate a 'split-point' in 'a' according to class 'c' i.e. 1 and 2. I wanna group 'a' into two class. Let say after calculation the split point is 8, then output will be {1,2,3,6,7,8} and {9,10} for 'a' means 1,2,3,6,7 and 8 will become {1,2,3,6,7,8}Sorry, but your description, coupled with the data you supplied, makes no sense to me.
How are you supposed to "split" column a according to 'c', when the values 7 and 8 appear for both 'c' values?
More description please.
My problem is: Which way will be easier to do this? Using what kind of input? (excel, txt, csv) ?I'd start with .txt. Mind you, I still don't understand the requirements...
Winston