binning data in equally spaced intervals

I would like to bin the data in 8 equally spaced bins. I have a the data accessible in vectors [X= conc, Y= alt] this what I have so far to create the bins

    binedge = linspace(min(alt),max(alt),6)

then I want to take the average alt and conc of each each bin and plot it. Thanks

[per isakson]

histc

[shobhit mehrotra]

can you elaborate?

 histc(.....)

thanks

[per isakson]

The description and examples of the documentation histc, Histogram bin counts (not recommended; use histcounts) are better than mine.

histcounts was introduced in R2014b. I've never used it.

[shobhit mehrotra]

How would i take the average of each bin?

Answer 1

[Josh Meyer]

As others noted, histcounts was introduced in R2014b and provides a great deal more flexibility for problems like this.

To use 8 bins, just do:

% Assume the first column is X, second column is Y
data = [rand(100,1), rand(100,1)];

% Find the bin placement based on the X values
[N,edges,bins] = histcounts(data(:,1),8);

The third output, bins, describes the bin placement of each element. So this makes finding the average X and Y value in each bin simple.

for n = 1:8
  bin_means(:,n) = mean(data(bins==n,:))';
end

bin_means =

    0.0751    0.1979    0.3342    0.4658    0.5691    0.7113    0.8691    0.9676
    0.5100    0.6264    0.4949    0.5172    0.5323    0.5556    0.4381    0.6514

[Stephen23]

Using some array preallocation would make this a more universal solution.

[Andrew Hurford]

This worked well for me - but for 1 result in a bin, mean defaults to selecting data across the data, so set the direction flag to 1

[Rahul Kesarkar]

How would it work if i have to bin data for a particular date range?

For example range : - weekday 12/06/2017 20:00 to 13/06/2017 08:00 and repeat this for every weekday. Weekend 9/06/2017 20:00 to 12/06/2017 08:00. Sum up all the values in the bin.

Answer 2

[Stephen23]

You can use histc to get bin the data, and then accumarray to get the mean of all of the values in each bin:

>> X = 0.25:0.5:5
X =
    0.25  0.75  1.25  1.75  2.25  2.75  3.25  3.75  4.25  4.75

>> [B,idx] = histc(X,0:5);
>> V = accumarray(idx(:),X,[],@mean)
V =
    0.5000
    1.5000
    2.5000
    3.5000
    4.5000

If you have a newer version, then you should use histcounts instead:

[B,~,idx] = histcounts(...);

[shobhit mehrotra]

I also want the X-value average for each bin, in addition to the y value. Thanks

[Stephen23]

You can simply repeat the accumarray call for each of your X and Y data vectors:

meanofXvalues = accumarray(idx(:),Xvalues,[],@mean)
meanofYvalues = accumarray(idx(:),Yvalues,[],@mean)