lsquares_estimates_exact SciMax Toolbox lsquares_residual_mse

SciMax Toolbox >> lsquares_mse

lsquares_mse

Maxima Function

Calling Sequence

lsquares_mse (D, x, e)

Description

Returns the mean square error (MSE), a summation expression, for the equation e in the variables x, with data D.

The MSE is defined as:

n
====
\                        2
>    (lhs(e ) - rhs(e ))
/           i         i
====
i = 1
--------------------------
           n

where n is the number of data and e[i] is the equation e evaluated with the variables in x assigned values from the i-th datum, D[i].

load(lsquares) loads this function.

Example:

(%i1) load (lsquares)$
(%i2) M : matrix ([1, 1, 1], [3/2, 1, 2], [9/4, 2, 1], [3, 2, 2], [2, 2, 1]);
                           [ 1  1  1 ]
                           [         ]
                           [ 3       ]
                           [ -  1  2 ]
                           [ 2       ]
                           [         ]
(%o2)                      [ 9       ]
                           [ -  2  1 ]
                           [ 4       ]
                           [         ]
                           [ 3  2  2 ]
                           [         ]
                           [ 2  2  1 ]
(%i3) mse : lsquares_mse (M, [z, x, y], (z + D)^2 = A*x + B*y + C);
           5
          ====
          \                 2                         2
           >    ((D + M    )  - C - M     B - M     A)
          /            i, 1          i, 3      i, 2
          ====
          i = 1
(%o3)     ---------------------------------------------
                                5
(%i4) diff (mse, D);
         5
        ====
        \                             2
      4  >    (D + M    ) ((D + M    )  - C - M     B - M     A)
        /           i, 1         i, 1          i, 3      i, 2
        ====
        i = 1
(%o4) ----------------------------------------------------------
                                  5
(%i5) ''mse, nouns;
               2                 2         9 2               2
(%o5) (((D + 3)  - C - 2 B - 2 A)  + ((D + -)  - C - B - 2 A)
                                           4
           2               2         3 2               2
 + ((D + 2)  - C - B - 2 A)  + ((D + -)  - C - 2 B - A)
                                     2
           2             2
 + ((D + 1)  - C - B - A) )/5
lsquares_estimates_exact SciMax Toolbox lsquares_residual_mse