# Find three largest and smallest values in column of table

10 views (last 30 days)
xander fong on 4 Aug 2015
Answered: Peter Perkins on 5 Aug 2015
Hello, so I have a 1500x15 table, in which one column represents profits for that asset. I would like to know how to code to return the three largest and smallest values in column 4. Additionally, I would like the code to also return the corresponding name (in column 1) of these values.

Peter Perkins on 5 Aug 2015
As Walter said, this is just sorting. In R2013b or later, use a table:
>> asset = {'a'; 'a'; 'a'; 'b'; 'b'; 'b'; 'c'; 'd'; 'e'};
>> price = rand(9,1);
>> x = randn(9,1);
>> data = table(asset,price,x)
data =
asset price x
_____ _______ _________
'a' 0.81472 2.7694
'a' 0.90579 -1.3499
'a' 0.12699 3.0349
'b' 0.91338 0.7254
'b' 0.63236 -0.063055
'b' 0.09754 0.71474
'c' 0.2785 -0.20497
'd' 0.54688 -0.12414
'e' 0.95751 1.4897
>> data = sortrows(data,'price');
>> smallest = data([1 2 3],:)
smallest =
asset price x
_____ _______ ________
'b' 0.09754 0.71474
'a' 0.12699 3.0349
'c' 0.2785 -0.20497
>> largest = data([end-2 end-1 end],:)
largest =
asset price x
_____ _______ _______
'a' 0.90579 -1.3499
'b' 0.91338 0.7254
'e' 0.95751 1.4897

Walter Roberson on 4 Aug 2015
Extract the values. sort() them, also returning the index as the second output of sort(). The first three sorted values are the three smallest values. The last three sorted values are the three largest values. The corresponding entries from the second output of sort can be used to index the name column.