Looking up the results from the function "LARGE"

K

Kirk

Hello,

I have an array that I'm using LARGE to find the top ten values. I want to
use VLOOKUP, or what ever else suits, on these top ten values to find other
values in the array. This works fine until the top ten that LARGE yields has
2 or more values the same. How would I eliminate a value that has already
been returned/looked up in order to find the next in line?

Thanks for your time?
 
E

Eli Shichrur

sort the data and insert new field that grow in 1 only if the astimated data
grows
use vlookup from 1 to 10 on the new fied
 
M

Max

Kirk said:
I have an array that I'm using LARGE to find the top ten values. I want to
use VLOOKUP, or what ever else suits, on these top ten values to find other
values in the array. This works fine until the top ten that LARGE yields has
2 or more values the same. How would I eliminate a value that has already
been returned/looked up in order to find the next in line?

One way ..
Try this sample from my archives collection (full details inside):
http://savefile.com/files/5659668
Simple Auto-Sort Example.xls
(Auto-sort table in descending order)
 
K

Kirk

Thanks people, but i'm using large to find to the top ten of 5 different KPIs
which i then plan to graph along with the value i'm looking up. At the
moment I'm sorting which is what i'm trying to avoid because the graphs
change each time i sort.
 
M

Max

The sample illustrates using an arb. tiebreaker in the criteria col which
enables a full descending sort of the entire source table to be extracted
elsewhere, inclusive ties / multiple ties, if any occurs. You could then
point to (& graph) or read off the top "10" in the extract. Think you could
also point your downstream VLOOKUP or INDEX/MATCH formulas at another key col
with unique values in the extract, eg KPI#?, KPI description?, etc instead of
the scores/points col used in the descending sort.
 
K

Kirk

Ok, Thanks for your help Max

Max said:
The sample illustrates using an arb. tiebreaker in the criteria col which
enables a full descending sort of the entire source table to be extracted
elsewhere, inclusive ties / multiple ties, if any occurs. You could then
point to (& graph) or read off the top "10" in the extract. Think you could
also point your downstream VLOOKUP or INDEX/MATCH formulas at another key col
with unique values in the extract, eg KPI#?, KPI description?, etc instead of
the scores/points col used in the descending sort.
 
Top