Why do image coordinate switches when using getpts function?

I have a 96x256x256 data, and I try to pick a pixel from the 256x256 to get data. When I use the getpts function, I get different x and y coordinates. Here is my approach to my data: 1. I have data that contain x and y coordinates, and data set(size of 96) at x-y coordinate. 2. I re-edited the data so I can put it into a matrix called tissue_map by doing a for loop:

for i = 1:length(xcoord)
    tissue_map(:,xcoord(i),ycoord(i)) = tissueData(:,i);
end

After this, I use getpts to get a coordinate on one slice of the matrix, say tissue_map(1,120,130):

fig = imagesc(tissue_map(1,:,:));
[x, y] = getpts(fig);
where I get:
x = 130
y = 120

What is it that I'm doing wrong? I couldn't figure out why.

Answer 1

[Walter Roberson]

I do not see a problem?

In matrix indexing, "up/down" is row indexing, the first index. Up/down corresponds to Y coordinates. So x = 130, y = 120 is row 120 column 130, index (120,130)

[Arbol]

I'm sorry, I think I put that in wrong. Let's say I try to get [x y] = [95 141]. I would get x = 95, y = 141 using getpts. It is correct coordinate wise, but if I use tissue_map(1,95,141) I would get different data set. I would have to use tissue_map(1,141,95) to get the right data set.

I guess what I'm trying to ask is, since my for loop is:

 tissue_map(:,xcoord(i),ycoord(i)) = tissueData(:,i);

why wouldn't it just return the right data set at first, by putting [x y] = [95,141]? why do I have to flip the coordinate to tissue_map(1,141,95), rather than using tissue_map(1,95,141)?

I keep running to this problem with the coordinate switching up. I tried to keep it straight with my coding. Do you have any suggestion?

[Walter Roberson]

How are you constructing xcoord and ycoord ?

[Arbol]

1. I have raw data that give x and y coordinate with size of 1x(256*256). Where each row of x increases by 1, each column of y increases by 1. I noticed that this is opposite than MATLAB's indexing. So:

2. I do some data processing to get my region of interest (xvec and yvec) to then put it in:

fit_xcoord = zeros(256,256);
fit_ycoord = zeros(256,256);
fit_intensity = zeros(length(time),256,256);
for i=1:length(xvec)
    IPx = xvec(i); IPy = yvec(i);
    fit_zcoord = zcoord;
    %-- Flip x and y coordinates
    fit_xcoord(IPx,IPy)= ycoord(IPx,IPy); 
    fit_ycoord(IPx,IPy)= xcoord(IPx,IPy);
    fit_intensity(:,IPx,IPy) = intensity_subtracted(:,IPx,IPy);
end

Where I flipped the xvec and yvec in the fit_xcoord and fit_ycoord so the indexing are similar to MATLAB's.

3. I use what I asked above to get my coordinate of interest by using getpts:

for i = 1:length(xcoord)
    tissue_map(:,xcoord(i),ycoord(i)) = tissueData(:,i);
end

fig = imagesc(tissue_map(1,:,:));
[x, y] = getpts(fig);
where I get:
x = 130
y = 120

Where xcoord = fit_xcoord and ycoord = fit_ycoord from step 2.