Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Library: metrics
See also: seq

Quantlet: sur
Description: Estimates a seemingly unrelated regression system by feasible generalized least squares

Usage: {b,s}=sur(surlist1,surlist2)
Input:
surlist1 list of matrices. If there are M equations then surlist1 will be a list of M+1 matrices. The first matrix is comprised of the vectors of observations of the left-hand side variables. Each of the next M matrices consists of the observations of the right-hand-side variables of one of the M equations of the system.
surlist2 list of string vectors. If there are M equations then surlist2 will be a list of M+1 string (column) vectors. Each vector contains the names of the variables of the corresponding matrix from surlist1.
Output:
b vector, containing the generalized least squares coefficient estimates
s vector, containing the estimated standard errors of the components of b

Note:

Example:
library("metrics")
z=read("carc.dat")
price=z[,2]
weight=z[,9]
foreign=(z[,14].<>1)
mpg=z[,3]
displ=z[,12]
length=z[,10]
n=rows(z)
one = matrix(n,1)
lhs=price~mpg~displ
x1=one~foreign~weight~length
x2=one~foreign~weight
x3=one~foreign~weight
surlist1=list(lhs,x1,x2,x3)
; setting up the string vectors for surlist2
yl="price"|"mpg"|"displ"
zl1="one"|"foreign"|"weight"|"length"
zl2="one"|"foreign"|"weight"
zl3="one"|"foreign"|"weight"
surlist2=list(yl,zl1,zl2,zl3)
{b,s}=sur(surlist1,surlist2)

Result:
You will find the following output in the output window

Contents of outputgls


[ 1,] "====================================================="
[ 2,] "   Generalized Least-squares estimates"
[ 3,] "====================================================="
[ 4,] " EQ  dep. var.      R2 "
[ 5,] "====================================================="
[ 6,] " 1   price                  0.586     "
[ 7,] " 2   mpg                    0.686     "
[ 8,] " 3   displ                  0.822     "
[ 9,] "====================================================="
[10,] "VARIABLE    Coef.Est.        Std.Err.          t"
[11,] "--------------------------------------------------"
[12,] "one         7635.062       3698.595          2.064"
[13,] "forei       3486.077        587.354          5.935"
[14,] "weigh          6.426          0.956          6.724"
[15,] "lengt       -116.055         32.896         -3.528"
[16,] "---------------------------------------------------------
[17,] "one           41.471          2.005         20.683"
[18,] "forei         -1.425          1.007         -1.414"
[19,] "weigh         -0.007          0.001        -11.093"
[20,] "---------------------------------------------------------
[21,] "one          -86.899         23.972         -3.625"
[22,] "forei        -25.257         12.043         -2.097"
[23,] "weigh          0.097          0.007         13.704"
[24,] "====================================================="



Author: A. Werwartz, 20000711
(C) MD*TECH Method and Data Technologies, 05.02.2006