Select the correct data from matrix

Hello. I have the following problem: For a product I predict the prices it will have the next day (24 prices). I use 5 models to predict. So I have 5 different predictions. I put these predictions in a matrix A (1X120). Next, I calcuate the error resulting between the acual values and the predicted values of each model and place the results in matrix B (24X5). Then, I select the smallest error for each one hour out of 24 and I put the result in matric C(1X24). What I am trying to do is in another matrix (for example D) to set the prediciton values that correspond to each error from matrix C. But beacuse the errors are different evry time I run the code, how can I find the corresponding values from matrix A? Your help is important!!!!

Answer 1

[Rik]

You're probably using min to select the smallest value, so you should use the second output to find the index of those values. Then you can use those col indices to find the linear indices in A.

[stelios loizidis]

Correctly, I used min. But I did not understand what you told me. Is it possible to write the commands?

[Rik]

Have you read the documentation of the min function? That will explain what the second output parameter does.

If you want me to edit your code you should post it.

[stelios loizidis]

I have read the documentation of min function. However, what I am trying to do is to have a separate matrix for each minimum error in matrix C, to have the corresponding value provided by matrix A. Below I put a piece of code

A=[]; % 1X120: predicted values
B=[]; % there errors:Actual values-Predicted Valeus
for i=1:24
    C(i)=min(B(i,:)) % In matrix C there are the small error from each row of matrix B
end

[Rik]

You can use the second output argument of min to find the index that contains the minimum value. If you know how the positions in B relate to the positions in A, you can use that to translate those indices in B to indices in A.