Description Usage Arguments Details Value See Also Examples
Calculates the maximum adjusted profile likelihood estimates.
1 |
fitted |
any fitted model object for which the maximum adjusted profile likelihood estimates can be calculated. |
... |
absorbs any additional argument. |
This function is generic (see methods
); method
functions can be written to handle specific classes of data. Classes
which already have methods for this function include: nlreg
.
the maximum adjusted profile likelihood estimates for all parameters of a regression model or for a subset of them.
mpl.nlreg
, nlreg.object
,
methods
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | data(metsulfuron)
metsulfuron.nl <-
nlreg( formula = log(area) ~ log( b1+(b2-b1) / (1+(dose/b4)^b3) ),
weights = ~ ( 1+dose^exp(g) )^2, data = metsulfuron, hoa = TRUE,
start = c(b1 = 138, b2 = 2470, b3 = 2, b4 = 0.07, g = log(0.3)) )
mpl( metsulfuron.nl, trace = TRUE )
##
options( object.size = 10000000 )
data(chlorsulfuron)
chlorsulfuron.nl <-
nlreg( log(area) ~ log( b1+(b2-b1) / (1+(dose/b4)^b3) ),
weights = ~ ( 1+k*dose^g*(b2-b1)^2/(1+(dose/b4)^b3)^4*b3^2*dose^(2*b3-2)/
b4^(2*b3)/(b1+(b2-b1)/(1+(dose/b4)^b3))^2 ),
start = c(b1 = 2.2, b2 = 1700, b3 = 2.8, b4 = 0.28, g = 2.7, k = 1),
data = chlorsulfuron, hoa = TRUE, trace = TRUE,
control = list(x.tol = 10^-12, rel.tol = 10^-12, step.min = 10^-12) )
mpl( chlorsulfuron.nl, trace = TRUE )
|
differentiating mean and variance function -- may take a while
iteration 1 : modified profile log likelihood = 0.3964742
iteration 2 : modified profile log likelihood = 0.3964742
Formula:
log(area) ~ log(b1 + (b2 - b1)/(1 + (dose/b4)^b3))
Variance function:
~(1 + dose^exp(g))^2 * exp(logs)
Higher order method used: Skovgaard's r*
Variance parameters
MMPLE MLE
g -1.274 -1.259
logs -3.716 -3.820
Regression coefficients
MMPLE MLE
b1 1.389e+02 1.389e+02
b2 2.471e+03 2.471e+03
b3 1.709e+00 1.709e+00
b4 7.727e-02 7.727e-02
Total number of observations: 40
Total number of parameters: 6
-2*Log Lmp -0.7929
Algorithm converged in 2 iterations
iteration 1 : log likelihood = -35.12366
iteration 2 : log likelihood = -35.10621
iteration 3 : log likelihood = -35.10528
iteration 4 : log likelihood = -35.10515
iteration 5 : log likelihood = -35.10513
iteration 6 : log likelihood = -35.10513
iteration 7 : log likelihood = -35.10513
iteration 8 : log likelihood = -35.10513
iteration 9 : log likelihood = -35.10513
iteration 10 : log likelihood = -35.10513
differentiating mean and variance function -- may take a while
iteration 1 : modified profile log likelihood = -35.92606
iteration 2 : modified profile log likelihood = -35.92892
iteration 3 : modified profile log likelihood = -35.92967
iteration 4 : modified profile log likelihood = -35.92967
Formula:
log(area) ~ log(b1 + (b2 - b1)/(1 + (dose/b4)^b3))
Variance function:
~(1 + k * dose^g * (b2 - b1)^2/(1 + (dose/b4)^b3)^4 * b3^2 *
dose^(2 * b3 - 2)/b4^(2 * b3)/(b1 + (b2 - b1)/(1 + (dose/b4)^b3))^2) *
exp(logs)
Higher order method used: Skovgaard's r*
Variance parameters
MMPLE MLE
g 2.645 2.603
k 1.081 1.010
logs -1.825 -1.888
Regression coefficients
MMPLE MLE
b1 2.2318 2.2063
b2 1657.3476 1657.3485
b3 2.8582 2.8426
b4 0.2752 0.2758
Total number of observations: 51
Total number of parameters: 7
-2*Log Lmp 71.86
Algorithm converged in 4 iterations
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.