Services & Resources / Wolfram Forums / MathGroup Archive
-----

MathGroup Archive 2008

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Function Programming Problems


Murray Eisenberg wrote:
> Two comments:
> 
> The use of a function name instead of a functional expression works just 
> as nicely for a user-defined function as for a built-in function.  For 
> example:
> 
>    f[x_] := x^3 Exp[-x]
>    LinearApproximation[f,0][x]
> 
> Second, since the original poster is writing the LinearApproximation 
> function for use in a calculus class, presumably this comes well before 
> the notion of series expansion or Taylor polynomials are are ever 
> discussed.  So it would be unfair to the students at this point to 
> "spoil" things by prematurely introducing Series.  Let them deal with 
> the special case of best local linear approximation, probably very early 
> in Calculus I, and some time later with best local quadratic 
> approximation.  Then they'll have something upon which to build the 
> generalization to best n-th degree polynomial local approximation.

Then why not let them find the linear and quadratic approximations for 
themselves, instead of providing a function to do it? My students would 
do the following.

f[x_] = x^2 Cos[x]

a = \[Pi]

p1[x_] = f[a] + f'[a] (x - a)

p2[x_] = f[a] + f'[a] (x - a) + f''[a]/2 (x - a)^2

Plot[{f[x], p1[x]}, {x, 0, 2 \[Pi]}]

Plot[{f[x], p2[x]}, {x, 0, 2 \[Pi]}]



-- 
Helen Read
University of Vermont


  • Prev by Date: Re: ParametricPlot precision problem
  • Next by Date: Re: question about sorting lists
  • Previous by thread: Re: Re: Function Programming Problems
  • Next by thread: Re: Function Programming Problems