For loop overwriting data

Hello,

I have this code below:

clear all;
names= {'AB_Clean','AK_Clean'};

data=[];
for iSubj=1:numel(names)
    filetoload= names{iSubj};
    t = readtable([filetoload '.xls']);
    data.id = num2str (iSubj);
    
    response = mod(t.MouseClickAngleToCenter,180);
    condition = mod(t.Condition,180);
    target = mod(t.CenterPackmanOriantation,180);
    
    for i=1:length(response)
        A=str2num(t.SurroundingPackmansRotations_tl_bl_tr_br_{i});
        if ~isempty(A)
            distractors_local(i,:) = mod(A,180);
        else
            distractors_local(i,:) = [nan nan nan nan];
        end
    end
    distractors_global = mod(t.PackmansRectangleRotation+90,180); %Original analysis
    %distractors_global = mod(t.PackmansRectangleRotation,180);%New collinearity analysis
    errors = circdist(response,target);
    errors(errors>90) = errors(errors>90) -180;
    errors(errors<-90) = errors(errors<-90) +180;
    distractors1 = circdist(distractors_global,target);
    distractors4 = circdist(distractors_local,target);
    distractors1(distractors1>90) = distractors1(distractors1>90) -180;
    distractors1(distractors1<-90) = distractors1(distractors1<-90) +180;
    distractors4(distractors4>90) = distractors4(distractors4>90) -180;
    distractors4(distractors4<-90) = distractors4(distractors4<-90) +180;
    distractors4 = distractors4(:,1:2);
    
    for c=unique(condition)'
    data.errors = errors(condition==c)';
    data.distractors = [distractors1(condition==c,:)'; distractors4(condition==c,[1 2])'];
    end
end

I want the loop to go through the data of each participant (i.e AB_Clean and AK_Clean) and save the data.errors, data.distractors and data.id variables for each participant in the "data" struct. There is something missing in my code, since the variables are being overwritten every time the loop starts. Any help with this would be very much appreciated.

Thanks!

-Mikel

Answer 1

[Walter Roberson]

clear all;
names= {'AB_Clean','AK_Clean'};

data = struct();

for iSubj=1:numel(names)
    filetoload= names{iSubj};
    t = readtable([filetoload '.xls']);
    data(iSubj).id = num2str(iSubj);
    
    response = mod(t.MouseClickAngleToCenter,180);
    condition = mod(t.Condition,180);
    target = mod(t.CenterPackmanOriantation,180);
    
    for i=1:length(response)
        A=str2num(t.SurroundingPackmansRotations_tl_bl_tr_br_{i});
        if ~isempty(A)
            distractors_local(i,:) = mod(A,180);
        else
            distractors_local(i,:) = [nan nan nan nan];
        end
    end
    distractors_global = mod(t.PackmansRectangleRotation+90,180); %Original analysis
    %distractors_global = mod(t.PackmansRectangleRotation,180);%New collinearity analysis
    errors = circdist(response,target);
    errors(errors>90) = errors(errors>90) -180;
    errors(errors<-90) = errors(errors<-90) +180;
    distractors1 = circdist(distractors_global,target);
    distractors4 = circdist(distractors_local,target);
    distractors1(distractors1>90) = distractors1(distractors1>90) -180;
    distractors1(distractors1<-90) = distractors1(distractors1<-90) +180;
    distractors4(distractors4>90) = distractors4(distractors4>90) -180;
    distractors4(distractors4<-90) = distractors4(distractors4<-90) +180;
    distractors4 = distractors4(:,1:2);
    
    
    ucond = unique(condition).';
    data(iSubj).conditions = ucond;
    ncond = length(ucond);
    data(iSubj).errors = cell(ncond,1);
    data(iSubj).distractors = cell(ncond,1);
    for cidx = 1 : ncond
      c = ucond(cidx);
      mask = condition==c;
      data(iSubj).errors{cidx} = errors(mask)';
      data(iSubj).distractors{cidx} = [distractors1(mask,:)'; distractors4(mask,[1 2])'];
    end
end

Now data will be a struct array with the same length as the number of names.

There is a new field conditions to record the unique conditions.

The errors and distractors fields have been turned into cell arrays, with one entry for each unique condition for this subject.

I do this instead of just concatenating everything together because there are a variable number of matches for each condition, and if you were to concatenate them all together you would not be able to tell which entry matched to which condition.

[Mikel Jimenez]

Dear Walter,

Thank you very very much for your help. Now I'll try to understand what's going on in the new part of the code:)

Thanks,

Mikel