|
View previous topic :: View next topic
|
| Author |
Message |
Dingu
New User
Joined: 06 Jul 2007
Posts: 10
Location: Chentrappinni
|
|
|
|
'Decile' is often used in the statistic world. It is a way of Ranking a list of Data; after grouping them into 10 different groups based on a key value.
A concept of Percentile.
If the key value is available in the database, then we have the TOP..PERCENT method in SELECT SQL which divides the selected rows into 10 groups and fetch the TOP n'th group. But here, the key value has to be calculated in a COBOL program and then sort the list of rows(or array fields) according to the key value and then divide the list or array into 10 group and rank the top group as 1, second as 2 and so on....
Any idea on a keyword similar to SQL TOP..PERCENT method in COBOL? Is there an easier method? |
|
| Back to top |
|
 |
CICS Guy
Senior Member
Joined: 18 Jul 2007
Posts: 2146
Location: At my coffee table
|
|
|
|
| Dingu wrote: |
| Any idea on a keyword similar to SQL TOP..PERCENT method in COBOL? Is there an easier method? |
Can't think of any, but you would not need to sort, just identify the highest (available/possible?) score and the lowest (available/possible?) score and divide that range into ten ranges....... |
|
| Back to top |
|
|
Dingu
New User
Joined: 06 Jul 2007
Posts: 10
Location: Chentrappinni
|
|
|
|
| CICS Guy wrote: |
| Dingu wrote: |
| Any idea on a keyword similar to SQL TOP..PERCENT method in COBOL? Is there an easier method? |
Can't think of any, but you would not need to sort, just identify the highest (available/possible?) score and the lowest (available/possible?) score and divide that range into ten ranges....... |
That's indeed a good suggestion.. i shall do the normal coding and test until i find any related COBOL keyword or method for the implementation to be different. |
|
| Back to top |
|
|
|
|
