MathGroup Archive 2000

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

Search the Archive

Re: N-Dimensional line

  • To: mathgroup at smc.vnet.net
  • Subject: [mg22798] Re: [mg22785] N-Dimensional line
  • From: BobHanlon at aol.com
  • Date: Sun, 26 Mar 2000 02:58:48 -0500 (EST)
  • Sender: owner-wri-mathgroup at wolfram.com

The built-in function Fit does what you want.

Let the dimension of the space be

dim = 5;

The variables to define the space are

vars = ToExpression[Table["x" <> ToString[k], {k, dim}]];

The functions to define the line are

funcs = Join[{1}, vars];

To generate test data

coef = 10*Table[Random[], {dim + 1}];

Clear[f];

f[x_List] := First[coef] + (Rest[coef].x) /; Length[x] == dim

data = Join[#, {f[#]}] & /@ Table[Random[], {dim + 2}, {dim}];

To Fit a line to the generated test data

Fit[data, funcs, vars]

3.9635834515677497 + 4.0699930252986904*x1 + 
  5.430269807608436*x2 + 5.302060419354726*x3 + 
  8.430356457003278*x4 + 1.4150730899130353*x5

Verifying that this is the same line as that used to generate the test data

% - f[vars] // Chop

0

Bob Hanlon

In a message dated 3/25/2000 5:11:54 AM, birdy00 at bu.edu writes:

>Does anyone have a reference and/or Mathematica code for finding the
>line
>in N-dimensional space that minimizes the distances from this line to a
>set of
>points in this space?
>


  • Prev by Date: Error in Symbolic Line Integral
  • Next by Date: ADMIN: Newsgroup and Mailing List off-line for three days
  • Previous by thread: Re: N-Dimensional line
  • Next by thread: Re: Re: N-Dimensional line