Re: writing functions that accept arrays/(matrcies?)

[David Grundberg]

Christopher Harvey wrote:
> Hello list,
> I'm pretty new to octave and I have a few questions. 
>
> I'll try to keep them in list form as much as possible.
>
> Are ranges, arrays and matrices the same thing in octave? (ie the same data 
> type?)
>   

Ranges (e.g. 1:4), cellarrays ({}) and matrices ([1 2 3]) are all 
arrays.  As an example, all arrays can be indexed with (), which returns 
an array of the same type. But cellarrays have a different class from 
matrices, so they store different things.

Note that ranges and matrices both return "double" from the class() 
function, so they are compatible in every way.  If you use typeinfo(), 
you can see what internal representation Octave chosen. typeinfo() is 
more for helping C++ developers debug than it is useful for m-file 
developers.

> If I write a function that accepts arrays (ranges?) like sin([1:10]) how can I 
> check to make sure that it's not a matrix that's actually passed in?
> How do other built in functions handle this in octave?
>   

You cannot make a difference between ranges and double matrices, except 
if you cheat and use typeinfo().  But I consider that bad style, because 
ranges and matrices belong to the same class.

If you have a lot of parameters and want to have error checking, you can 
use is*() functions, which checks the class. That way you are not 
dependent on the internal representation in Octave.

> I would like to write a function that can accept and produce arrays, can I get 
> a template that can do this? ie: myFunction([1:10], 1, 2, 3)
> that returns [x1, x2, x3....... x10]
>   

Something like this

function y = linearfunc(x, k, m)
 y = x .* k + m;
endfunction

> thanks for clearing this up,
> Chris.
> _______________________________________________
> Help-octave mailing list
> address@hidden
> https://www-old.cae.wisc.edu/mailman/listinfo/help-octave
>