Sort and accumulate data in a matrix

Hi everyone,

I have written a code but it's to messy, I'm sure one could sum it up.

This is the data I have:

data= [0	-50	
-10	-50	
-10	-50	
0	-40	
-10	-40	
10	-40	
10	-30	
-10	-30	
0	-30	
... and so on till +50.

The output I want should be seperated for the three values in the first column so that the output for 0, 10 and -10 should separatley look like this:

[ -50     0     
   -40     0     
   -30     0     
   -20     1     
   -10     1     
     0      2     
    10     6    
    20     6    
    30     6     
    40     6    
    50     6   ]

so thst the first column lists every value once and the second column the amount of occurance of ones for that specific value.

My code:

%%0 (delete others)
data(find(data(:,1)==10),:)=[]
data(find(data(:,1)==-10),:)=[]

%% 10 (delete others)
data(find(data(:,1)==0),:)=[]
data(find(data(:,1)==-10),:)=[]

%%  -10 (delete others)
data(find(data(:,1)==10),:)=[]
data(find(data(:,1)==0),:)=[]

And then for every data (unfortenately) separately the following:

for i = size(data)
    data1=sum(data(1:3,2))
    data2=sum(data(4:6,2))
    data3=sum(data(7:9,2))
    data4=sum(data(10:12,2))
...    
end
comp_data=[data1, data2, data3, data4]'
test_angle=unique(data(:,2));
dataAll=horzcat( test_angle,comp_data)

I am pretty sure one could sum it up by different for/if loops so that I don't have to run it for 0, 10 and -10 separately and independetly from fixed row values.

Thanks a lot!!!

[Walter Roberson]

data(find(data(:,1)==10),:)=[]
data(find(data(:,1)==-10),:)=[]
%% 10 (delete others)
data(find(data(:,1)==0),:)=[]
data(find(data(:,1)==-10),:)=[]

?? The -10 are already gone -- you deleted them in the second statement.

Have you considered

data(ismember(data(:,1), [-10 0 10]), :) = [];

[Elli]

Your suggestion leads leads to an empty struct!

And you are right, I am sorry, it was a typo.

%% 10 (delete others)

data(find(data(:,1)==0),:)=[]

data(find(data(:,1)==-10),:)=[]

Answer 1

[Stephan]

Here is an example:

A = [-10 0 10 10 20 -20 30 -30 40 -50 50 -40 -10 0 0 10 20 -20 30 -30 40 -50 50 -40 40 40 40];
groups = findgroups(A);
occurences = splitapply(@numel,A,groups);
result = [unique(A); occurences].'

[Guillaume]

Note that unique(A) is the 2nd output of findgroups.

Simpler code overall:

[groups, values] = findgroups(A(:));  %can also be done with [values, ~, groups] = unique(A);
result = [values, accumarray(groups, 1)]

or:

values = unique(A);
result = [values; histcounts(A, [values, Inf])].'

[Elli]

Hi Stephan, thank you!

This counts the occurance of the values in the first column (A), not the corresponding values in the second row!

[Stephan]

@Elli This was just to give you a direction how to improveyour approach. Together with the commentfrom Guillaume you should be able to do this in a proper way.

@Guillaume Thank you for the hint, again a nice new fact i did not notice so far.