Error using trace in matrices manipulation

Hello I am trying to compute 2 different matrixes of various size using the code below and I am getting an error

in the line :

field2 = 'accuracy';
value2 = (2*trace(value1)+sum(sum(2*value1)))/(numOfClasses*totalSamples);

it is showing the error

Error using trace (line 12)

Matrix must be square.

Error in confusionmatStats (line 46)

field2 = 'accuracy'; value2 = (2*trace(value1)+sum(sum(2*value1)))/(numOfClasses*totalSamples);

Please your help will be appreciated

Thanks

Tino

  T =  
    5.1000    3.5000    1.4000    0.2000
    6.3453    3.2126    4.3695    1.4453
    5.4421    3.3916    1.8178    0.2547
    5.6874    2.8253    4.3822    1.2537
    5.7000    4.3327    1.6263    0.4379
    5.6042    2.9631    3.5116    1.2621
t = 
    5.1000    3.8000    1.9000    0.4000
    6.3453    2.9379    4.1990    1.2621
    4.3884    2.9958    1.2347    0.1505
    5.3537    3.0000    4.3569    1.4411

using the code below:

nargin = any number ;
if nargin < 1
    value1 = T;
else
    value1 = confusionmat(T, t);
end
numOfClasses = size(value1,1);
totalSamples = sum(sum(value1));    
field2 = 'accuracy';  value2 = (2*trace(value1)+sum(sum(2*value1)))/(numOfClasses*totalSamples);
[TP,TN,FP,FN,sensitivity,specificity,precision,f_score] = deal(zeros(numOfClasses,1));
for class = 1:numOfClasses
   TP(class) = value1(class,class);
   tempMat = value1;
   tempMat(:,class) = []; % remove column
   tempMat(class,:) = []; % remove row
   TN(class) = sum(sum(tempMat));
   FP(class) = sum(value1(:,class))-TP(class);
   FN(class) = sum(value1(class,:))-TP(class);
end

for class = 1:numOfClasses
    sensitivity(class) = TP(class) / (TP(class) + FN(class));
    specificity(class) = TN(class) / (FP(class) + TN(class));
    precision(class) = TP(class) / (TP(class) + FP(class));
    f_score(class) = 2*TP(class)/(2*TP(class) + FP(class) + FN(class));
end

field3 = 'sensitivity';  value3 = sensitivity;
field4 = 'specificity';  value4 = specificity;
field5 = 'precision';  value5 = precision;
field6 = 'recall';  value6 = sensitivity;
field7 = 'Fscore';  value7 = f_score;
stats = struct(field1,value1,field2,value2,field3,value3,field4,value4,field5,value5,field6,value6,field7,value7);

Answer 1

[Jan]

The error message is clear:

  T = [ ...
    5.1000    3.5000    1.4000    0.2000
    6.3453    3.2126    4.3695    1.4453
    5.4421    3.3916    1.8178    0.2547
    5.6874    2.8253    4.3822    1.2537
    5.7000    4.3327    1.6263    0.4379
    5.6042    2.9631    3.5116    1.2621];
t = [ ...
    5.1000    3.8000    1.9000    0.4000
    6.3453    2.9379    4.1990    1.2621
    4.3884    2.9958    1.2347    0.1505
    5.3537    3.0000    4.3569    1.4411];

size(T)
size(confusionmat(T, t))

trace requires a square matrix as input. So which sizes do you get?

Please note, that it is unlikely that readers can guess broken code without any further details. Seeing the failing code only does not allow to guess its intention.

[Tino]

Error using trace (line 12)

Matrix must be square.

Error in zondo (line 168)

field2 = 'accuracy'; value2 = (2* (trace(value1)+sum(sum(2*value1))))/(numOfClasses*totalSamples);

Find the errors I am getting above

actually I know there are of different matrices what other way do I mingle around that

Please help

thanks

[Tino]

I get a (6,4) matrix

[Jan]

@Tino: You provide a non-square matrix to the trace command. This cannot work, because a trace is defined for a square matrix only. The error message tells this clearly already. We cannot guess, what you want to do instead, because the current code is not meaningul in a mathematical sense.

"actually I know there are of different matrices what other way do I mingle around that" - I do not understand, what this means and cannot know, which formula you want to implement.