Using histogram() data without plotting or calculating the figure.

Hi, this is my first post here.
I really need to optimize a function so as to make it as fast as possible. In it, I use the histogram function. All I need is the histogram().Values and histogram().BinEdges data. I don't need to plot anything and I need to keep the computation time to the bare minimum. The code I use is roughly:
fig=figure; set(fig,'visible','off');
H=histogram(...);
h=H.Values;
b=H.BinEdges;
% Some data treatment
%%%%
close(fig);
By doing 'tic-tocs' I found out that these precise lines were taking up 90% of the execution time. Is there any way to have access to the histogram data without opening a figure, or any other way of optimizing time ? I mean not opening a figure at all, not just hiding it, as I've already done that.
Thanks in advance !

1 Comment

Possibly a different way to compute a histogram without plotting a figure is to use the function histcounts to compute the histogram counts and edges. Later the histogram can then be plotted with the histogram function.
E.g.:
[counts, edges] = histcounts(data, 100) % Computes the histogram counts and edges for 'data' without plotting
...
histogram('BinEdges', edges, 'BinCounts', counts) % Plots a histogram with edges 'edges' and counts 'counts'.

Sign in to comment.

 Accepted Answer

If you only want the histogram values and not the plot, use the histcounts (or histc) functions.

5 Comments

Great ! It was exactly the data I needed. It divided the computation time by 10 on the histogram, and that of my whole program by more than 2. Thank you !
But these functions do not provide the edges :(
The second output from histcounts contains the edges.
Thanks Star
I have the same problem with my codes efficiency
But even with histcount the code is still slow

Sign in to comment.

More Answers (0)

Categories

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!