Anybody know how to simplify this?

I have two pieces of code that I would like to know if there is another way to write them

for i = 1:m
  for j = 1:n
     H(j,i) = some_function(W(i,:),X(j,:)); 
  end
end

and

for i = 1:n
  Y(i,:) = other_function(Y(i,:));
end

Answer 1

[Walter Roberson]

If you have the stats toolbox, then

H = pdist2(W, X, @somefunction)

For the other one, I cannot think of any way to simplify the code. There are other ways of writing it, such as

Y = cell2mat( arrayfun(@(i) other_function(Y(i,:)), (1:size(Y,1)).', 'uniform', 0) );

but I would by no means say that is simpler or more efficient.

There is a special case that can be written more simply: if Y is a table or timetable, then https://www.mathworks.com/help/matlab/ref/rowfun.html

Y = rowfun(@otherfunction, Y);

Answer 2

[Jacob Ward]

Not sure about the first one, but for the second one, the for loop and indexing is unnecessary. See this example:

This...

>> M = [0 1 2; 3 4 5; 6 7 8];
>> for i = 1:3
>>     M(i,:) = sin(M(i,:));
>> end

M =
         0    0.8415    0.9093
    0.1411   -0.7568   -0.9589
   -0.2794    0.6570    0.9894

Yields the same results as...

>> M2 = [0 1 2; 3 4 5; 6 7 8];
>> M2 = sin(M2);

M2 =
         0    0.8415    0.9093
    0.1411   -0.7568   -0.9589
   -0.2794    0.6570    0.9894