Get started for free
Visual Studio Developer Center >
> 
> 
FS0003 when overriding operators

Beantwortet FS0003 when overriding operators

  • Tuesday, March 19, 2013 7:53 PM
     
      Has Code
    Sign In to Vote
    1

    Hello. I'm writing an optimization program using Rosenbrock methods - http://en.wikipedia.org/wiki/Rosenbrock_methods

    So i'm trying to describe relations between scalar and vector, so i'm writing a code like this

    [<EntryPoint>]
    let main argv = 
        let ( *? ) (a:float, vector:float[]) = 
            vector |> Array.map (fun x->x * a)
        let (+?) (a:float, vector:float[]) = 
            vector |> Array.map (fun x->x + a)
        let (+%) (X:float[], Y:float[]) = 
            Array.map2(fun x y -> x + y) X Y        
        let n = 2
        let d = Array.init n (function i -> (Array.init n (function index -> if index = i then 1. else 0.)))
        let x : float[] = Array.zeroCreate n
        let y = x
        let k, j = 1, 1
        let f (x : float[]) = x.[0]**2.
        let fu lambda = f(y +% (lambda  *? d.[1]))      
        printfn "%A" argv
        0

    but there's an error FS0003: 

    This value is not a function and cannot be applied

    in

    let fu lambda = f(y +% (lambda  *? d.[1]))

    how to fix it?




    • Edited by PsilonRus Tuesday, March 19, 2013 8:12 PM
    • Edited by PsilonRus Tuesday, March 19, 2013 8:13 PM
    • Edited by PsilonRus Tuesday, March 19, 2013 8:14 PM
    •  
     

All Replies

  • Wednesday, March 20, 2013 12:27 AM
     
     Answered Has Code
    Sign In to Vote
    2

    When you create custom infix operators you probably should use curried form of function definition:

    let ( *? ) (a:float) (vector:float[]) =         
    vector |> Array.map (fun x->x * a)    
let ( +? ) (a:float) (vector:float[]) =         
    vector |> Array.map (fun x->x + a)    
let ( +% ) (X:float[]) (Y:float[]) =         
    Array.map2(fun x y -> x + y) X Y
    Otherwise compiler doesn't know how to call infix operator with tupled input.

    Petr

    • Marked As Answer by PsilonRus Wednesday, March 20, 2013 5:03 AM
    •  
     
  • Wednesday, March 27, 2013 11:49 PM
     
     
    Sign In to Vote
    0

    Hi PsilonRus and Petr Plepilov.

    This may interest you:
    http://fsharpnews.blogspot.com/2011/01/gradient-descent.html

    I'm working on "Refolding Planar Polygons" by Iben, O'Brien and Demaine.

    Climbing the learning curve of optimization.

    And the state of .NET maths ...
    FSharp.Math -> Math.Net -> Math.Numerics -> Cloud Numerics
    I'm not sure I got that right ...

    Solvers too.

    I'm interested in your progress.

    Thanks. PLH OOE Art

    Peace Love Happiness PLH Only One Earth OOE