How to get mean of all variables in a table?

Hello, A table ('kub') that I am trying to analyse has a dimension of 1882x33. Its a mixed table with double and strings values. The 1st and 2nd variable is contains strings. Therefore, I used the following line of code to get the mean of all Variables except 1st and 2nd.

func = @mean;
B = varfun(func,kub{:,3:end});

However, I am getting the following error.

Check for missing argument or incorrect argument data type in call to function 'varfun'.+

Can anyone help me?

Answer 1

[Steven Lord]

The varfun function requires one of its inputs to be a table or timetable array (or a tall table or timetable.)

which -all varfun

/MATLAB/toolbox/matlab/datatypes/tabular/@tabular/varfun.m  % tabular method
/MATLAB/toolbox/matlab/bigdata/@tall/varfun.m               % Shadowed tall method

In your call the first input is a function handle. When you index into a table or timetable array using curly braces what you receive is not itself a table or timetable.

T = array2table(magic(4))

T = 4×4 table
    Var1    Var2    Var3    Var4
    ____    ____    ____    ____

     16       2       3      13 
      5      11      10       8 
      9       7       6      12 
      4      14      15       1 

smallerTable = T(2:3, 2:4) % table array

smallerTable = 2×3 table
    Var2    Var3    Var4
    ____    ____    ____

     11      10       8 
      7       6      12 

doublePiece = T{2:3, 2:4} % double array

doublePiece = 2×3

11 10 8 7 6 12

If you want to compute the mean of all the elements of the doublePiece variable don't use varfun.

M1 = mean(doublePiece, "all")

M1 = 9

M2 = mean(doublePiece, 1)

M2 = 1×3

9 8 10

M3 = mean(doublePiece, 2)

M3 = 2×1

9.6667 8.3333

But if you want to operate on just the numeric variables in your table, you can use a function handle to select as the input variables only those that are numeric.

T.stringvar = ["a"; "b"; "c"; "d"]

T = 4×5 table
    Var1    Var2    Var3    Var4    stringvar
    ____    ____    ____    ____    _________

     16       2       3      13        "a"   
      5      11      10       8        "b"   
      9       7       6      12        "c"   
      4      14      15       1        "d"   

varfun(@mean, T, 'InputVariables', @isnumeric)

ans = 1×4 table
    mean_Var1    mean_Var2    mean_Var3    mean_Var4
    _________    _________    _________    _________

       8.5          8.5          8.5          8.5   

[Kazi Alam]

varfun(@mean, T, 'InputVariables', @isnumeric)

It worked like a charm. Thank you @Steven Lord.