Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

5-point Gaussian Quadrature using constructed approximant in Matlab

  1. Mar 30, 2008 #1
    1. The problem statement, all variables and given/known data
    6.3.b highlighted in attachment.
    Have solved part a (which gives the approximant used in part b) and problem 3.8 (which gives the original function). 3.8 was definitely solved correctly. Part a could be wrong, but the solution seems OK.
    a = acreage
    y = yield
    from 3.8 - p1,p2,c1,c2 are prices and consumption in periods 1 and 2.
    2. Relevant equations
    a = 0.5+0.5*Ey(f(ay)) where Ey is the expected price in terms of y and f(ay)=p1(s).

    3. The attempt at a solution
    function [fval] = f63b(y)
    fval = 0.5+0.5*Exp;

    [y,w] = qnwlogn(20,0,0.01);
    Exp = w'*p1;
    fprintf('Exp'); disp(Exp)

    a = broyden('f63b',s/2);
    fprintf('a'); disp(a)

    Gives value for expectation that is accurate, but then the following error:
    ??? Undefined function or variable 'Exp'.
    Cannot find solution for a! Should we have 5 nodes or 20? When change to qnwlogn(5,0,0/01) get no value for Exp.
    Last edited: Mar 30, 2008
  2. jcsd
  3. Mar 30, 2008 #2
    Need to evaluate function [p1hat] = p1hat(a,y) using routine funeval(coef,fspace1,s), where s=ay. Matlab gives the following error:

    If you get this message, you have not correctly installed the CompEcon Toolbox
    Please read the README file for installation details, epsecially the section concerning MEX files
    ??? Error using ==> CEtools/private/chebeval at 22

    But toolbox is installed correctly!
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook