2 Dataset Name: Thurber (Thurber.dat)
\r
5 Starting Values (lines 41 to 47)
\r
6 Certified Values (lines 41 to 52)
\r
7 Data (lines 61 to 97)
\r
9 Procedure: Nonlinear Least Squares Regression
\r
11 Description: These data are the result of a NIST study involving
\r
12 semiconductor electron mobility. The response
\r
13 variable is a measure of electron mobility, and the
\r
14 predictor variable is the natural log of the density.
\r
17 Reference: Thurber, R., NIST (197?).
\r
18 Semiconductor electron mobility modeling.
\r
25 Data: 1 Response Variable (y = electron mobility)
\r
26 1 Predictor Variable (x = log[density])
\r
28 Higher Level of Difficulty
\r
31 Model: Rational Class (cubic/cubic)
\r
32 7 Parameters (b1 to b7)
\r
34 y = (b1 + b2*x + b3*x**2 + b4*x**3) /
\r
35 (1 + b5*x + b6*x**2 + b7*x**3) + e
\r
38 Starting Values Certified Values
\r
40 Start 1 Start 2 Parameter Standard Deviation
\r
41 b1 = 1000 1300 1.2881396800E+03 4.6647963344E+00
\r
42 b2 = 1000 1500 1.4910792535E+03 3.9571156086E+01
\r
43 b3 = 400 500 5.8323836877E+02 2.8698696102E+01
\r
44 b4 = 40 75 7.5416644291E+01 5.5675370270E+00
\r
45 b5 = 0.7 1 9.6629502864E-01 3.1333340687E-02
\r
46 b6 = 0.3 0.4 3.9797285797E-01 1.4984928198E-02
\r
47 b7 = 0.03 0.05 4.9727297349E-02 6.5842344623E-03
\r
49 Residual Sum of Squares: 5.6427082397E+03
\r
50 Residual Standard Deviation: 1.3714600784E+01
\r
51 Degrees of Freedom: 30
\r
52 Number of Observations: 37
\r