I am trying to find the difference in two columns on my csv file to plot them and my current code isnt working.

I have a code that intakes all of my participants per there csv file and outputs there data by the objects they have seen however I cannot code the difference between the known midpoint of the object versus what they have answered. Am I overlooking something? I did change some of the info to keep more discreet sorry. Thank you to anyone who helps!

% Select CSV file and read data
[filename, path] = uigetfile('*.csv', 'Select The CSV to View Subject Data');
myCSV = readtable(fullfile(path, filename));

% Define the column indices for KeyPressCount and ResponseTime
Midpoint = 60;
Final_Answered_Midpoint = 76;

%Assigning condition names by category {row, column}
ConditionNames{1,1} = '1';
ConditionNames{2,1} = '2';
ConditionNames{3,1} = '3';
ConditionNames{4,1} = '4';
ConditionNames{5,1} = '5';
ConditionNames{6,1} = '6';
ConditionNames{7,1} = '7';
ConditionNames{8,1} = '8';
ConditionNames{9,1} = '9';
ConditionNames{10,1} = '10';
ConditionNames{11,1} = '11';
ConditionNames{12,1} = '12';

NbrCond = size(ConditionNames,1);

%Difference between midpoint and final slider position that was answered
difference = myCSV{:,76} - myCSV{:,60};

%Creating a loop that through each of the Categories (ConditionNames) in column 1 of myCSV and
%reports the row indices every time each category is found
for condi = 1:NbrCond
    ConditionIndex(condi).RowIndices = [];
    Indices = strfind(table2array(myCSV(:,1)), ConditionNames{condi,1}); %Strfind the current (condi) condition in the array 
    counter = 0;
    for j = 1:size(Indices,1)
        if ~isempty(Indices{j,1})
            counter = counter + 1;
            ConditionIndex(condi).RowIndices(counter) = j;
        end
    end
end


%Loops over each category (ConditionNames) in CSV to find data (ex:
%KeyPressCount) and stores it in AllData 
for condi = 1:NbrCond
    ConditionInstances = size(ConditionIndex(condi).RowIndices,2);
    counter = 0;
    for Condi_instance_i = 1:ConditionInstances
        CurrentRowIndex = ConditionIndex(condi).RowIndices(Condi_instance_i);
        counter = counter + 1;
        %AllData(condi).Category(counter,1) = final_slider_position_pix-Midpoint;
        AllData(condi).Category(counter,1) = myCSV{CurrentRowIndex, difference};
    end
end

% Calculating Averages, Standard Deviations, and Standard Errors and
% storing them in AllData by Category (Row)
for condi = 1:NbrCond
    AllData(condi).Mean = mean(AllData(condi).Category);
    AllData(condi).StdDev = std(AllData(condi).Category);
    AllData(condi).StdErr = AllData(condi).StdDev / sqrt(size(AllData(condi).Category, 1)); 
end


CategoryAvgs = cellfun(@(x) mean(x), {AllData(:).Category});
CategoryStdDevs = cellfun(@(x) std(x), {AllData(:).Category});
CategoryStdErrs = CategoryStdDevs ./ sqrt(cellfun(@numel, {AllData(:).Category}));

%Creating variable to store all of the means and standarderrors for
%plotting purposes
Means = [AllData.Mean];
StdErrs = [AllData.StdErr];

%Creating a figure to summarize results
figure;
bar([AllData.Mean]);
hold on; 
errorbar(1:NbrCond, CategoryAvgs, CategoryStdErrs, '.k'); 

ylim([-10 25])
xlabel('Tools');
ylabel('Difference');

% Set x-axis labels (goes in order of AllData.Mean)
xticklabels({'1', '2', '3', '4', ...
    '5', '6', '7', '8', ...
    '9', '10', '11', '12'});
xtickangle(45)
hold off;

[dpb]

Please format your code and attach at least a small data file...

[dpb]

Just some comments on coding efficiency in MATLAB...

NbrCond = 12;                               % set the number (or read from somewhere else)
ConditionNames=string(1:NbrCond).'          % create the name automagically by code, not manually

ConditionNames = 12x1 string array
    "1"
    "2"
    "3"
    "4"
    "5"
    "6"
    "7"
    "8"
    "9"
    "10"
    "11"
    "12"

Making into a categorical array wouldn't be bad or just leaving them as numeric instead of strings would make comparison easier without needing the string matching at all.

% Select CSV file and read data
[filename, path] = uigetfile('*.csv', 'Select The CSV to View Subject Data');
myCSV = readtable(fullfile(path, filename));

% Define the column indices for KeyPressCount and ResponseTime
Midpoint = 60;
Final_Answered_Midpoint = 76;
...
%Difference between midpoint and final slider position that was answered
difference = myCSV{:,76} - myCSV{:,60};

You read as a table, the above could be just

difference=myCSV.Var76-myCSV.Var60;

assuming you didn't have a row of other variable names. Ideally, those columns would also be identified with usable and recognizable variable names.

%Creating a loop that through each of the Categories (ConditionNames) in column 1 of myCSV and
%reports the row indices every time each category is found
for condi = 1:NbrCond
    ConditionIndex(condi).RowIndices = [];
    Indices = strfind(table2array(myCSV(:,1)), ConditionNames{condi,1}); %Strfind the current (condi) condition in the array 
    ...

Same sort of thing; you have a table, use table referencing of the desired variables instead of turning the whole thing into an array every pass through a loop.

If the conditions were numeric or categorical, the above could be more like

for condi = 1:NbrCond
  Indices=matches(myCSV.Var1,ConditionNames(condi));  % for string comparison
  ...

It's not clear exactly what you're shooting for in the next; give a problem description of what the end result is to be and undoubtedly somebody will be along and deal with the logic.

More than likely, any statistics or calculations by such a subsetting can be done with varfun with grouping variables without explicit looping.

Answer 1

[Shishir Reddy]

Hi Kathryn

As per my understanding, you are facing issues in your code while trying to find the difference between two columns in a csv file. This is because of the wrong syntax used in the following code snippet.

for condi = 1:NbrCond
    ConditionInstances = size(ConditionIndex(condi).RowIndices,2);
    counter = 0;
    for Condi_instance_i = 1:ConditionInstances
        CurrentRowIndex = ConditionIndex(condi).RowIndices(Condi_instance_i);
        counter = counter + 1;
        AllData(condi).Category(counter,1) = myCSV{CurrentRowIndex, difference};
    end
end

The line - AllData(condi).Category(counter,1) = myCSV{CurrentRowIndex, difference}; is incorrect because difference is a vector of differences, not a column index.

The correct approach is to use the difference vector directly, indexing it with CurrentRowIndex.

AllData(condi).Category(counter,1) = difference(CurrentRowIndex);

This ensures that the Category field in AllData is populated with the correct difference values for each condition.

I hope this helps.

[Kathryn]

Thank you so much ShiShir, this helped alot!!!