| Line 62... |
Line 62... |
| 62 |
observation \eqn{i} is an average of \eqn{w_i} observations, which is
|
62 |
observation \eqn{i} is an average of \eqn{w_i} observations, which is
|
| 63 |
natural for binomials specified as proportions but less so for a
|
63 |
natural for binomials specified as proportions but less so for a
|
| 64 |
Poisson, for which prior weights are ignored with a warning.
|
64 |
Poisson, for which prior weights are ignored with a warning.
|
| 65 |
|
65 |
|
| 66 |
For a gamma GLM the shape parameter is estimated by maximum likelihood
|
66 |
For a gamma GLM the shape parameter is estimated by maximum likelihood
|
| 67 |
(using function \code{\link[MASS:gamma.shape.glm]{gamma.shape}} in package
|
67 |
(using function \code{\link[MASS]{gamma.shape}} in package
|
| 68 |
\CRANpkg{MASS}). The interpretation of weights is as multipliers to a
|
68 |
\CRANpkg{MASS}). The interpretation of weights is as multipliers to a
|
| 69 |
basic shape parameter, since dispersion is inversely proportional to
|
69 |
basic shape parameter, since dispersion is inversely proportional to
|
| 70 |
shape.
|
70 |
shape.
|
| 71 |
|
71 |
|
| 72 |
For an inverse gaussian GLM the model assumed is
|
72 |
For an inverse gaussian GLM the model assumed is
|