Plotting in 3D: Probability Density Plots Over Time
14 views (last 30 days)
There are many ways of plotting in 3D in matlab. I have got some reasonable results using the following method. This post is to enquire how other people might tackle the common problem of representing PDFs moving over time.
N = 1E4; %number of experiments, eg Monte Carlo
T = 3E3; %number of timestamps
wHat= randn(T,N); %at each point in time you get a distribution of probabilities.
ksData = NaN(T,N);
binVals = linspace(0,1, N); %assign the probabilities to one of these buckets
for t = 1: T
ksData(t,:) = ksdensity(wHat(t,:), binVals);
surf(1:T, binVals, ksData');
set(gca, 'XLIM', [0 T]);
set(gcf, 'color', 'white');
This seems to give a decent visual representation of whats going on.
However, I am aware that I haven't smoothed across time, only across experiments.
A related post is here: http://mathworks.co.uk/matlabcentral/newsreader/view_thread/319825