mata function and loading data from excel

[Charles Cadestin]

Hello,
This is one of my first attempt to create a mata function... And this is also my first post here.
I have several xlsx files with data from 2000 to 2014. I am doing some mata operations from these different sets. I am stuck at the very beginning of my code... As I cannot create a function that returns the matrix I want:

mata: mata drop test()
mata
real matrix test(scalar year)  //the output will be a matrix so I type "real matrix"
{
real matrix Z 
real scalar N, K, I, C
N = 42
K = 20
I = 2
C = 5
b=xl() // excel file loaded tool
b.load_book("2014_m2.xlsx") // i identify the workbook from where i want to import 
b.set_sheet("sheet1") // i identify the spreadsheet
Z=b.get_number((4,4+N*K*I-1),(4,4+N*K*I-1)) //import the matrix
return(Z)
}
end
mata: test(2014)

However, the code returns the following error:

type mismatch: exp.exp: transmorphic found where struct expected
r(3000);

just after b.load_book("2014_m2.xlsx")

: real matrix test(scalar year) 
> {
> real matrix Z
> real scalar N, K, I, C
> N = 42
> K = 20
> I = 2
> C = 5
> b=xl()
> b.load_book("2014_m2.xlsx")
type mismatch:  exp.exp:  transmorphic found where struct expected
r(3000);

I do not understand what I am doing wrong here... When I try to directly import the data, it works perfectly but as soon as I introduce these peaces of code in a matrix... Error! I would appreciate a lot some guidances.

Thank you very much,

Charles.

[Sergio Correia]

I get that error often, it took me a while to understand...

Anyways, you are doing two things wrong:

1. make your second line "mata:" instead of "mata". Note the colon!!!

2. You are declaring every variable except for b. Add a declaration for b and it should work:

class xl scalar b

(The error occurs because when b is not declared, stata assumes is of type transmorphic. But then when you say b.load_book it thinks it's a structure (not even a class) and complains. I get why the error occurs, and it kinda sorta makes sense, but still annoying)

[Charles Cadestin]

Dear Sergio,
I just tested it and indeed it worked perfectly well. Thank you very much.

Charles