Get maximum values across a cell array with multiple indexing.

Perhaps the title of my question isn't phrased at best, but hopefully the description will provide clear insight. I have the following couple of arrays:

idListReq = {'stock.A', 'stock.A', 'stock.A', 'stock.B', 'stock.B', 'stock.B', 'stock.B'}; 
factorData = [1, 4, 2, 9, 10, 1, 3];

Each element of idListReq has its numerical assignment in factorData. I would like to get the maximum across the groups of values that are tied to a single ID (e.g. values for 'stock.A' are 1, 4 and 2, meaning the max will be 4).

The desired outcome should like this:

idList = {'stock.A', 'stock.B'}; 
outcome = [4, 10];

What I used at the beginning was cellfun, and it worked, but I want to know if there's a better way for this, without using loops.

fData = cellfun(@(x) max(factorData(strcmp(idListReq, x))), idList);

Best regards and thanks in advance!

Answer 1

[dpb]

Use grouping variables -- there's still a looping construct inside, of course, but you don't have to write it explicitly...

>> groupsummary(factorData.',idListReq.',"max")
ans =
          4.00
         10.00
>>

NB: 1. Must use column vectors, hence the ." transpose operator.

2. Looks like the id list variable would be ideal candidate for a categorical variable

[Stephen23]

The second output argument is probably also useful:

idListReq = {'stock.A', 'stock.A', 'stock.A', 'stock.B', 'stock.B', 'stock.B', 'stock.B'}; 
factorData = [1, 4, 2, 9, 10, 1, 3];
[V,C] = groupsummary(factorData(:),idListReq(:),@max)

V = 2×1

4 10

C = 2×1 cell array
    {'stock.A'}
    {'stock.B'}

[dpb]

Good point, @Stephen