To the best of my knowledge,I did not find any answers anywhere to the answer to my problem. I consider myself pretty decent in Matlab.
I have the trajectory of a tumor recorded vs time like on this image. I would like to compute cumulative distribution that would show the time vs displacement of the tumor from its ideal position at x=0 like on this picture generated with another software.
What the cumulative graph means is that we can find the total time outside a certain position the tumor spent during acquisition. You see that the position of the tumor at the position 0 is the length of the entire acquisition time(~300 seconds).If we are looking for the time the tumor spent outside 1.1 mm from their ideal position, it would indicate ~100 seconds.The time the tumor spent outside 2.8mm becomes really close to 0s.
Any code that would help me get such a would be great. I strongly sense that there is something to do with cumsum, cdf etc, but I really have not managed to find a proper function. My next option would be to bin it myself and write the code for it.
Thank you for you help.
You can find the distribution using hist(x). Then by calculating the backwards cumsum() you can draw the desired plot.
clc, clear all, close all
seconds = 303; % Amount of time that passed during the test
datapoints = 3000; % Amount of Datapoints in your vector
x = randn(datapoints,1);
[counts,centers] = hist(abs(x),sort([0;unique(abs(x))]));
sumX = sum(counts);
cumsumX = cumsum(counts);
time = (sumX - [0 cumsumX(1:end-1)])*seconds/datapoints; % Normalize result with factor
figure
plot(centers, time)