How to perform a scatter plot based on density in MATLAB?

Question

MathWorks Support Team 2017-10-31

回答： MathWorks Support Team 2018-1-3

I have two vectors x & y and I am looking for a way to create a scatter plot of x & y based on their degree of closeness or density in MATLAB.

Answer 1

The 'scatplot' command takes in column matrix x & y and performs a density based scatter plot as shown in this example:

>> x = randn(1,1000);
>> y = randn(1,1000);
>> scatplot(x,y);
>> colorbar;

The link to the 'scatplot' can be found here :

https://www.mathworks.com/matlabcentral/fileexchange/8577-scatplot

Another File Exchange function called 'dscatter' performs a similar plot.

The File Exchange link can be found here:

>> x = randn(1000,1);
>> y = randn(1000,1);
>> dscatter(x,y)

We can visualize the density within a grid in MATLAB by binning using the hist3 function and subsequently plotting it as follows:

>> x = randn(1,1000);
>> y = randn(1,1000);
>> n = hist3([x', y']);
>> pcolor(n);

To query the coordinates for a color range, please refer the following:

The function 'getDataForColorRange' in the attachment accepts an Axes handle returned by the 'dscatter' function and a color range.

For example, let us plot the following:

>> x = randn(1000,1);
>> y = randn(1000,1);
>> f = dscatter(x,y);
>> colorbar;

We can query all the X and Y coordinates that are present in the color range 0.7 to 0.9 as follows:

>> [x,y,idx] = getDataForColorRange(f,[0.7 0.9]);

To view this data, we can update the scatter plot as follows:

>> c = f.Children;
>> c.CData(~idx) = NaN;