Get the positions of the first logicals (multiple sets) in an array

Hi. I am trying to find where the value of an array increases to a particular threshold (thresh=5). For example, I started off with:

x = [0.5,1.6,1.5,5.2,8.5,9.3,6.3,2.2,1.9,1.8,7.4,6.4,6.5,2.3,2.4,8.5,7.6,5.3,2.0,2.4]

I converted this to a logical :

thresh = 5; 
x_log = x>thresh;

which gives me:

x_log = [0   0   0   1   1   1   1   0   0   0   1   1   1   0   0   1   1   1   0   0]

What I actually need is a logical that just has the first 1 of each group. So the output should be [0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0] ... is there a function I don't know of that can do this? I can't find it. Or do you suggest another way to do this using the original array?

Thanks a lot in advance!

[BobbyRoberts]

Long way I did this is:

x_idx = x>5;
x_idx = find(ischange(double(x_idx),'Threshold',0));
x_idx(2:2:end,:) = [];
x_new = zeros(length(x),1);
x_new(x_idx) = 1;

[Image Analyst]

Why do you say it works?

x = [0   0   0   1   1   1   1   0   0   0   1   1   1   0   0   1   1   1   0   0];
x_idx = x>5;
x_idx = find(ischange(double(x_idx),'Threshold',0));
x_idx(2:2:end,:) = [];
x_new = zeros(length(x),1);
x_new(x_idx) = 1

x_new = 20×1

0 0 0 0 0 0 0 0 0 0

No element of x_new is 1.

See my solution below in the official "Answers" section of this page.

[BobbyRoberts]

x = [0.5,1.6,1.5,5.2,8.5,9.3,6.3,2.2,1.9,1.8,7.4,6.4,6.5,2.3,2.4,8.5,7.6,5.3,2.0,2.4]

[BobbyRoberts]

if you use the original x array, sorry. Not the logical. I'm converting it to logical here using x_idx = x>5.

Answer 1

[Voss]

x = [0.5,1.6,1.5,5.2,8.5,9.3,6.3,2.2,1.9,1.8,7.4,6.4,6.5,2.3,2.4,8.5,7.6,5.3,2.0,2.4];
thresh = 5; 
x_log = x>thresh

x_log = 1×20 logical array
   0   0   0   1   1   1   1   0   0   0   1   1   1   0   0   1   1   1   0   0

One way to get the indices of the first 1s:

idx = find(diff([false x_log]) == 1)

idx = 1×3

4 11 16

Another way:

idx = strfind([false x_log],[false true])

idx = 1×3

4 11 16

(In either case, a false/0 element is prepended to x_log to correctly handle the case that x_log starts with a true/1.)

Then, regardless of how you get idx, build x_new from idx:

x_new = false(1,numel(x));
x_new(idx) = true

x_new = 1×20 logical array
   0   0   0   1   0   0   0   0   0   0   1   0   0   0   0   1   0   0   0   0

[BobbyRoberts]

nice! thank you. very similar to what i did, but more logical (pun maybe intended).

[Voss]

You're welcome!

Answer 2

[Image Analyst]

Try this:

x = [0.5,1.6,1.5,5.2,8.5,9.3,6.3,2.2,1.9,1.8,7.4,6.4,6.5,2.3,2.4,8.5,7.6,5.3,2.0,2.4];
mask = x > 5

mask = 1×20 logical array
   0   0   0   1   1   1   1   0   0   0   1   1   1   0   0   1   1   1   0   0

props = regionprops(mask, 'PixelIdxList');
for k = 1 : length(props)
    startingIndexes(k) = props(k).PixelIdxList(1);
end
% Show in command window:
startingIndexes

startingIndexes = 1×3

4 11 16

[BobbyRoberts]

thanks!