Organize matrix data into Bins using loop

Hi All,

I have some time series data in a matix, called NorthBlindChSpeed1. What I want to do is organise that data into evenly spaced Bins. The code I have so far, whcih works, is as shown below

NorthBlindChSpeed1 = [0.2; 0.4; 0.8; 0.8; 1; 1.19; 0.3; 1.41; 1.47; 1.51];

B1 =(NorthBlindChSpeed1(:, 1)< 0.2  & NorthBlindChSpeed1(:,1)>= 0);
B2 =(NorthBlindChSpeed1(:, 1)< 0.4  & NorthBlindChSpeed1(:,1)>= 0.2);
B3 =(NorthBlindChSpeed1(:, 1)< 0.6  & NorthBlindChSpeed1(:,1)>= 0.4);
B4 =(NorthBlindChSpeed1(:, 1)< 0.8  & NorthBlindChSpeed1(:,1)>= 0.6);
B5 =(NorthBlindChSpeed1(:, 1)< 1.0  & NorthBlindChSpeed1(:,1)>= 0.8);
B6 =(NorthBlindChSpeed1(:, 1)< 1.2  & NorthBlindChSpeed1(:,1)>= 1.0);
B7 =(NorthBlindChSpeed1(:, 1)< 1.4  & NorthBlindChSpeed1(:,1)>= 1.2);
B8 =(NorthBlindChSpeed1(:, 1)< 1.6  & NorthBlindChSpeed1(:,1)>= 1.4);

This code works and it assigns the values in the Matrix NorthBlindChSpeed1 into the appropriate bins. The trouble with this code is that one has to manually assign the data to each Bin. For the real data set I will require 250 Bins... Is there a more efficient way to Bin this data??? Perhaps using a For Loop.

I appreciate your help.

Answer 1

[Stephen23]

"Is there a more efficient way to Bin this data???"

Using inbuilt functions will be the most efficient approach, e.g. DISCRETIZE or HISTCOUNTS:

V = [0.2; 0.4; 0.8; 0.8; 1; 1.19; 0.3; 1.41; 1.47; 1.51];
E = 0:0.2:1.6; % bin edges
idy = discretize(V,E)

idy = 10×1

2 3 5 5 6 6 2 8 8 8

[cnt,~,idx] = histcounts(V,E)

cnt = 1×8

0 2 1 0 2 2 0 3

idx = 10×1

2 3 5 5 6 6 2 8 8 8

These will be much better approaches than copy-and-pasting lines of code, using loops, or numbering variable names.

[Ben Whitby]

Thanks, that works in terms of Binning the data; however, is there a way to access the data in each Bin. Say for example I want to go on and sum all of the values in Bin 8, as per the code below. It's not clear how I would do that using the binning method you have proposed.

NorthBlindChSpeed1 = [0.2; 0.4; 0.8; 0.8; 1; 1.19; 0.3; 1.41; 1.47; 1.51];
B1 =(NorthBlindChSpeed1(:, 1)< 0.2  & NorthBlindChSpeed1(:,1)>= 0);
B2 =(NorthBlindChSpeed1(:, 1)< 0.4  & NorthBlindChSpeed1(:,1)>= 0.2);
B3 =(NorthBlindChSpeed1(:, 1)< 0.6  & NorthBlindChSpeed1(:,1)>= 0.4);
B4 =(NorthBlindChSpeed1(:, 1)< 0.8  & NorthBlindChSpeed1(:,1)>= 0.6);
B5 =(NorthBlindChSpeed1(:, 1)< 1.0  & NorthBlindChSpeed1(:,1)>= 0.8);
B6 =(NorthBlindChSpeed1(:, 1)< 1.2  & NorthBlindChSpeed1(:,1)>= 1.0);
B7 =(NorthBlindChSpeed1(:, 1)< 1.4  & NorthBlindChSpeed1(:,1)>= 1.2);
B8 =(NorthBlindChSpeed1(:, 1)< 1.6  & NorthBlindChSpeed1(:,1)>= 1.4);

NorthBlindPowerB8 = sum(Power(B8,1));

[Stephen23]

"It's not clear how I would do that using the binning method you have proposed."

Using arrays, because this is MATLAB. Forget about writing out lots of numbered variable names like that, that is a dead-end. And copy-and-pasting lines of code is just doing the computer's job for it, best avoided.

You should use ACCUMARRAY or perhaps GROUPSUMMARY or something similar:

V = [0.2; 0.4; 0.8; 0.8; 1; 1.19; 0.3; 1.41; 1.47; 1.51];
P = rand(1,numel(V)); % corresponding power values
E = 0:0.2:1.6; % bin edges
[cnt,~,idx] = histcounts(V,E);
A = accumarray(idx,P(:),[8,1],@sum)

A = 8×1

0 0.8271 0.6048 0 1.2723 1.6261 0 1.8971

B = zeros(8,1);
B(cnt>0) = groupsummary(P(:),idx,@sum)

B = 8×1

0 0.8271 0.6048 0 1.2723 1.6261 0 1.8971

[Ben Whitby]

Thanks, The accumarray works well.