How do I build a matrix using two vectors?

Question

So I need to build a matrix of x and y coordinates. I have the x stored in one matrix called vx=0:6000; and y stored in Vy=repmat(300,1,6000);.

Values in x are 0,1,2,...,5999,6000. Values in y are 300,300,...,300,300.

How do I build a "vector" with the x,y coordinates above?

It would look like this [(0,300);(1,300);...;(5999,300);(6000,300)].

After I finish doing this, I am going to want to find the distance between another fixed point x,y (that I will replicate 6000 times) and the vector above, in order to make a distance graph over time.

Thank you so much!

Suever · Accepted Answer · 2016-10-10T20:22:05.510

4

You can just use horizontal concatenation with []

X = [Vx(:), Vy(:)];

If you want to compute the distance between another point and every point in this 2D array, you could do the following:

point = [10, 100];
distances = sqrt(sum(bsxfun(@minus, X, point).^2, 2));

If you have R2016b or newer you can simply do

distances = sqrt(sum((X - point).^2, 2));

edited Oct 10 '16 at 20:22

answered Oct 10 '16 at 19:30

Suever

64,497
14
82
101

@user3105028 You use `:` exactly as I have it written. That forces `x` and `y` to be column vectors. – Suever Oct 10 '16 at 19:38
>> p1 = [Vx(:), Vy(:)] Error using horzcat Dimensions of matrices being concatenated are not consistent. – belgarion Oct 10 '16 at 19:39
@user3105028 You want the last one you have written but you have 6001 elements in vx and 6000 in vy – Suever Oct 10 '16 at 19:40
Since R2016b you should be able to get rid of the `bsxfun` – Adriaan Oct 10 '16 at 20:11
What would you replace bsxfun with, Adriaan? – belgarion Oct 10 '16 at 20:16
@user3105028 Updated to include what he was referring to – Suever Oct 10 '16 at 20:22

score 3 · Answer 2 · answered Oct 10 '16 at 20:02

A slightly more elegant alternative (in my opinion) is the following:

Vx = (0:1:6000).';
C = [Vx 0*Vx+300];  % Just a trick to avoid the overly verbose `repmat`.

p = [10,100];       % Define some point of reference.
d = pdist2(C,p);    % The default "distance type" is 'euclidian' - which is what you need.

This uses the pdist2 function, introduced in MATLAB 2010a, and requires the Statistics and Machine Learning Toolbox.

How do I build a matrix using two vectors?

2 Answers2