How to make a matrix that descends with each row, but all columns are the same?

Hey, so I am trying to make a matrix that looks like this in general:

5 5 5 5 5

4 4 4 4 4

3 3 3 3 3

2 2 2 2 2

1 1 1 1 1

I vaguely remember from a class I took that I had to multiply a 1:5 matrix into a ones 5x5 matrix.. or something like that.

Could anyone help me figure out a way to get this type of matrix for any size? (for example, going from 255 down to 0 or something along those lines)

Thanks!

[Specifically, I need to make a matrix 500x500 that goes down from 255 to 0] So,

255 255 ... 255

254 254 ... 254

.

.

.

201 201 ... 201

etc.

Answer 1

[the cyclist]

One way:

N = 255;
M = 500;

vec = (N:-1:0)';

A = repmat(vec,1,M);

Answer 2

[Stephen23]

A simple matrix multiplication does the trick, no need for slow repmat:

>> (5:-1:1)' * ones(1,4)
ans =
   5   5   5   5
   4   4   4   4
   3   3   3   3
   2   2   2   2
   1   1   1   1

This is a neat trick using indexing:

>> X = (5:-1:1)';
>> X(:,ones(1,4))
ans =
   5   5   5   5
   4   4   4   4
   3   3   3   3
   2   2   2   2
   1   1   1   1

Both of these tend to be faster than repmat (100x100 matrix, 1000 iterations):

Elapsed time is 0.17901 seconds.   % matrix multiply
Elapsed time is 0.157009 seconds.  % indexing
Elapsed time is 1.37308 seconds.   % repmat

[the cyclist]

Huh. I get very different relative timings:

N =  99;
M = 100;

vec = (N:-1:0)';

tic
for ii = 1:1000
    A = vec*ones(1,M);
end
toc

tic
for ii = 1:1000
    B = vec(:,ones(1,M));
end
toc

tic
for ii = 1:1000
    C = repmat(vec,1,M);
end
toc

results in

Elapsed time is 0.014031 seconds.
Elapsed time is 0.007379 seconds.
Elapsed time is 0.004367 seconds.

and even more favorable results for repmat for larger arrays.