Discretizing a column array, and performing operations on elements of another column array of the same length that lie in the same bins as the original one

I have a data array that is composed of column vectors X, Y, & Z (coordinates) and an array F that has one or multiple columns of the same length. The number of rows of A is usually in the order of 100k to even a couple of million, so that I need to reduce this number to a fraction of that. Let's assume:

A=[X Y Z F]

X, Y & Z are column vectors of length i, and F is an array with i rows and j columns.

Using the discretize command,I would like to discretize column Z into a column with a lower number of rows (let's say going from i rows to only i/1000 rows), and I would like to calculate the average value of every individual coumn of F that lies in the same bin as the discretized Z.

Answer 1

[Matt J]

G=findgroups( discretize(Z,edges) );

Fbinned = splitapply(@(z) mean(z,1) , F, G);

Answer 2

[Catalytic]

I=findgroups(discretize(Z,edges)) ;

S=sparse(I,(1:numel(I))', 1);
S=S./sum(S,2);

Fbinned = S*F;