How can I color the area under a histfit curve?

Hello everyone, I am trying to make in Matlab a plot called heated density plot in R that is basically a smoothed histogram whose area under the curve is colored accordingly to the belonging to a certain "bin". Can everyone help me doing this? I would really appreciate this. I attach an example of what the outcome should look like more or less. Thank you

Answer 1

[Tala]

I assume you have an array that contains the envelope of your curve. If yes, this should help you! something like:

section1=1:0.1:2;

section2=2:0.1:3;

area(section1,yourarray);

hold all

area(section2,yourarray);

[Camilla Ancona]

Thank you for your answer! In truth I changed my mind and I am trying to visualize my data through a violin plot but i have the same issue to color the area of the violin or/and the datapoints color accordingly to their belonging of a certain class. here there is my attempt

I'm trying to visualize my data by means of this package, but I was wondering if there is a way to color differently the datapoints accordingly to their belonging of a certain range of equivalently the area of the violin. this is my attemot in substituiting the CData matrix with arranged RGB triplets but this approach does not works

vs = violinplot(data,[],'Width',0.2,'Bandwidth',0.3,'ViolinColor',[0 0 1],'ViolinAlpha',0.5,'EdgeColor',[0 0 1],'BoxColor',[0 0 0],'MedianColor',[1 0 0]);

hold on

h = vs(1).ScatterPlot;

cdata = vs.ScatterPlot.CData;

c = repmat(cdata,[1446 1]);

c(1:freq_new_net(1),:) = repmat([1 0 0],[freq_new_net(1) 1]);

c(freq_new_net(1)+1:freq_new_net(1)+freq_new_net(2),:) = repmat([1 1 0],[freq_new_net(2) 1]);

c(freq_new_net(1)+freq_new_net(2)+1:freq_new_net(1)+freq_new_net(2)+freq_new_net(3),:) = repmat([1 1 1],[freq_new_net(3) 1]);

c(freq_new_net(1)+freq_new_net(2)+freq_new_net(3)+1:freq_new_net(1)+ freq_new_net(2)+freq_new_net(3)+freq_new_net(4),:) = repmat([0 0 1],[freq_new_net(4) 1]);

c(freq_new_net(1)+freq_new_net(2)+freq_new_net(3)+freq_new_net(4)+1:1446,:) = repmat([0 1 0],[freq_new_net(5) 1]);

h.CData = c;

vs.ScatterPlot.CData = c;

Answer 2

[DGM]

I have no familiarity with violin plots or your data, and I don't know which violin plot tools you chose to use. If you're trying to essentially control the resolution of your colormap, consider a simplified example:

x = 1:100;
y = 1:100;
nbins = 5;

% find which bin each point belongs in
[~,~,idx] = histcounts(y,nbins); 
% start with a reduced colormap
cmap = jet(nbins); 

% combine to get cdata
scatter(x,y,[],cmap(idx,:))

I don't know if that's helpful in your case.