Controlchart function with the conditional variance?
1 view (last 30 days)
Show older comments
Karim Hasouna on 12 Sep 2020
Answered: Ayush Gupta on 17 Sep 2020
I am making a time series with the conditional variance and I want to create a plot.
This is my right code.
%% Model chart
v1 = sqrt(sigma2_hat)
ont = ones(1418);
UCP = on*(1.96)
LCP = on*(-1.96)
qv1 = UCP.*v1
h1 = plot(dates,y1,'Color','k')
h2 = plot(dates,qv1,'r--','LineWidth',2)
h3 = plot(dates,qv2,'r--','LineWidth',2)
Then, I want to determine the values that are out on the Upper and lower limits (red line).
I did this
st = controlchart(y1,'charttype','i','nsigma',2,'sigma',v1)
If I use a constant v1 (variance) works but I have a matrix and also my time series is a matrix with the same length.
Then, I should know how many points go beyond limits and create a legend.
Ayush Gupta on 17 Sep 2020
The problem that is arising here is because of the fact the argument given to sigma that is v1 is, it can’t take multiple output. So, a matrix or an array argument to sigma should not be given. A possible workaround here could be to use a for loop and give value of v1 at each iteration. Refer to the following code for the workaround
st = controlchart(y1,'charttype','i','nsigma',2,'sigma',v(1))
Find more on Half-Normal Distribution in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!Start Hunting!