Zach is correct about the direct answer to the question.
An interesting side note is that the following two loops do not execute the same:
for i=1:10000
% do something
end
for i=[1:10000]
% do something
end
The first loop creates a variable i that is a scalar and it iterates it like a C for loop. Note that if you modify i in the loop body, the modified value will be ignored, as Zach says. In the second case, Matlab creates a 10k-element array, then it walks all elements of the array.
What this means is that
for i=1:inf
% do something
end
works, but
for i=[1:inf]
% do something
end
does not (because this one would require allocating infinite memory). See Loren's blog for details.
Also note that you can iterate over cell arrays.
When iterating over cell arrays of strings, the loop variable (let's call it f) becomes a single-element cell array. Having to write f{1} everywhere gets tedious, and modifying the loop variable provides a clean workaround.
% This example transposes each field of a struct.
s.a = 1:3;
s.b = zeros(2,3);
s % a: [1 2 3]; b: [2x3 double]
for f = fieldnames(s)'
s.(f{1}) = s.(f{1})';
end
s % a: [3x1 double]; b: [3x2 double]
% Redefining f simplifies the indexing.
for f = fieldnames(s)'
f = f{1};
s.(f) = s.(f)';
end
s % back to a: [1 2 3]; b: [2x3 double]
I think this is what the OP really wants:
array = -1:0.1:10
for i=1:numel(array)
disp(array(i))
end