cell2mat has a very specific condition

hey all! so i've been playing with cells and structures etc.

i have a structure 'students' with fields:name,id,grades.

in grades i have a nx1 cell array where n is the number of grades per students. the grades are in string and not numbers.

i wanted to calculate every student's avarage. i used a for loop and cell2mat and it gave me an nx2 matrix of char objects.

the problem is the last students, one of his grades is 100 which is 3 letters instead of 2 and because of that cell2mat has no idea how to compensate for it.

i know i could have made things easier for myself if i made a more normal structure in the first place but i'm too far in the excercise now lol.

any idea how to work with this?

[DGM]

Show us what code you have and provide an example of your data so that we can see what's going on.

[daniel slama]

of course, my bad.

this is my attempt to calculate every student's avarage with notes of what i tried to to every line.

[a b]=size(students)
for i = [1:1:b]
     charmat=cell2mat(students(i).grades) %creates the nx2 matrix and gets stuck with the last student
    [c d]=size(charmat)
    nummat=[]
    for x = [1:1:c]
        temp=str2num(charmat(x,:) 
        nummat=[nummat ; temp] %makes a number matrix out of the char matrix
    end
students(i).avarage=mean(nummat)
end

Answer 1

[dpb]

Indeed, you made dereferencing the data about as hard as could by putting it into cell array -- a struct could/can hold an ordinary array; it shouldn't be too hard to change the code to make the grades arrays numeric to start with...but one way to your objective here is\

>> arrayfun(@(i)mean(str2double(students(i).grades)),1:numel(students),'UniformOutput',true)

ans =

   84.0000   98.5000   86.1667   57.3333   87.2500

>> 

[daniel slama]

this is incredible. i did mess up with the struct, guess i still need to work on that.

thank you so much!

[dpb]

Will become familiar idiom (along with the companion functions for cell arrays and sructure fields) with time...