matlab to R: function calling and @

Question

I use R but I am translating code from matlab to R. I have reached a section which I cannot grok.

My research shows that the @ allows you to call a function by another name with fixed variables e.g.

g = @(b) f(a1, b, c1)

allows me to call f only specifying b by doing g(b)

In the code I am working with there is a function

function dN = WW(N,h,A,P,aA,aP,bA,bP)

at some point in the code it appears WW is called, but is called with

f = @(t,N) WW(N,h,A,P,aA,aP,bA,bP)

Why I am so confused is that t,N are mentioned nowhere else in the code....but h,A,P,aA,aP,bA,bP are all defined prior.

does anyone recognise this structure and what might be going on?

Answer 1

You are correct in your assessment of @. @(t) is what is known as an anonymous function. @(t) will thus return a handle to a function that takes in one variable t. Basically, it's a function that takes in one parameter, t. The rest of the parameters are defined previously in your workspace.

As an example, your statement of g = @(b) f(a1, b, c1) allows you to encapsulate this function call into another function called g, and a1 and c1 are previously defined in your workspace. This function will thus rely on one variable that goes into the function, which is b. As such, these parameters will remain static when you call the function g, and you can change b on the fly. In other words, every time you call g, a1 and c1 will always stay the same while b will change depending on what you put into g (check out Lexical Scope). Obviously, should a1 and c1 change in your workspace, this behaviour will be reflected when you call g the next time as well.

Now, with the other function call. t is never used, but N is! As such, N will dynamically change when you vary it, but t will have no effect on that function handle of f. No matter how you vary t, the output of f will be the same, provided that you don't change N!

In any case, your assessment is indeed correct.

Answer 2

at some point in the code it appears

function dN = WW(N,h,A,P,aA,aP,bA,bP)

is called, but is called with

f = @(t,N) WW(N,h,A,P,aA,aP,bA,bP)

Why I am so confused is that t,N are mentioned nowhere else in the code....but h,A,P,aA,aP,bA,bP are all defined prior.

That's completely ok. What happens here is that if you do

f(value_for_t, value_for_N)

it calls

WW(value_for_N,h,A,P,aA,aP,bA,bP)

(oops? t isn't used...)

so everything which contains t or N is replaced or gone.