working with numbers in the array names in for/ if loop

Hi

I am struggling with a program in matlab where I need to make arrays with numbers varying in their name in the loop itself:

For example: given dataset with columns: data_column_1, data_column_2, data_column_3 . . . upto 10

Lets say:

………………………….

For i=1:10
array_(i)= mean(data_column_(i) w.r.t data_column_1 )     %( for all 10  arrays )
end

………………………………..

Solution must be like

array_1 has mean of data in data_column_1 wrt data_column_1

array_2 has mean of data in data_column_2 wrt data_column_1

array_3 has mean of data in data_column_3 wrt data_column_1 and so on.

I will really appreciate some solution here. Thanks

[Stephen23]

"I will really appreciate some solution here"

Use indexing.

https://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

Answer 1

[Dyuman Joshi]

Dynamically naming variables is not a good coding practice.

Read - TUTORIAL: Why Variables Should Not Be Named Dynamically (eval)

As you have numbers only, you should concatenate all the data in a numerical array and directly use mean with the specific dimension.

[Sonali]

Can you show how? Through an example?

[Dyuman Joshi]

How do you get the arrays data_column_1, data_column_2, ... data_column_n?

Are they obtained as output from a code? If so, could you share the code?

The basic idea is -

%I assume the arrays are a row vector
data_column = vertcat(data_column_1,  data_column_2, ..,  data_column_n);
out = mean(data,2)

If the arrays are a column vector, then use horzcat and use 1 instead of 2 for mean.

If you can provide more information about the questions I asked above, we might be able to optimize the code further.

[Sonali]

Heres the code. My dataset(arr) has 3 columns, I need to find mean of all columns for the unique values of column 1. then plot those averages wrt to unique values i.e plot(col1,col2), plot(col1,col3). In my prog, I am getting Multiple values of c(nx3 matrix), which is a variable to store the unique mean values for each iteration. I have 10 bins. And I get different c-values for each iteration. It gives me error that size of arrays are not same on both sides. so program is stuck here. I will appreciate ome help, if you can.Thanks

.......................................................

for i= 1:10
    if isempty( arr) )==0
        for j=1:3 
            c(:,j)=splitapply(@mean,arr(:,j),findgroups(arr(:,1)));     %<----------- error line.
        end
        c
        plot(log(c(:,2)),c(:,1),'b.')
        
        hold on
        
    end
end

[Dyuman Joshi]

From what I understood by the description you gave, try this -

(I didn't understand the bit about the 10 bins.)

idx = findgroups(arr(:,1));
u = unique(idx);
c = zeros(numel(u),3);
for j=1:3 
    c(:,j)=splitapply(@mean,arr(:,j),idx);
end

figure
plot(c(:,1), c(:,2), 'b.')
hold on
plot(c(:,1), c(:,3), 'r.')

If this does not work, please attach the data for "arr".

[Dyuman Joshi]

Any updates, @Sonali?