How can I plot outliers and original data together?
4 views (last 30 days)
Show older comments
I set a threshold for my outliers, a then find them by "any" command. Now I want to plot the original data and the outliers together. How can I do that?
Temp = data(:,2); % Separating the feature vector
Temp_z_score = (Temp - mean(Temp)) / std(Temp); % Calculation of the z-score (z-score = the number of standard deviations a point is below or over the mean)
threshold = abs(Temp_z_score)>1.8;
U = any(threshold,2);
0 Comments
Answers (1)
Cris LaPierre
on 15 Oct 2022
Edited: Cris LaPierre
on 15 Oct 2022
Wouldn't plotting the original data already contain the outliers? I assume you want to plot each group with a different linespec.
Since threshold is a logical array, use logical indexing when you plot. See the bottom of this page for an explanation.
Temp = rand(100,1);
Temp_z_score = (Temp - mean(Temp)) / std(Temp);
threshold = abs(Temp_z_score)>1.8;
plot(Temp(~threshold),'bo','DisplayName','Data')
hold on
plot(Temp(threshold),'r*','DisplayName','Outlier')
hold off
legend
2 Comments
Cris LaPierre
on 15 Oct 2022
You need to specify an x value then. You have not indicated there is one in your OP. If you don't have one, then x is the element number.
Temp = rand(100,1);
X=1:length(Temp);
Temp_z_score = (Temp - mean(Temp)) / std(Temp);
threshold = abs(Temp_z_score)>1.8;
plot(X(~threshold),Temp(~threshold),'bo','DisplayName','Data')
hold on
plot(X(threshold),Temp(threshold),'r*','DisplayName','Outlier')
hold off
legend
See Also
Categories
Find more on Data Distribution Plots 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!