Akaike weights
Calculate, extract or set normalized model likelihoods (‘Akaike weights’).
Weights(x) Weights(x) <- value
x |
a numeric vector of information criterion values such as AIC, or
objects returned by functions like |
value |
numeric, the new weights for the |
The replacement function can assign new weights to an "averaging"
object, affecting coefficient values and order of component models.
For the extractor, a numeric vector of normalized likelihoods.
On assigning new weights, the model order changes accordingly, so assigning
the same weights again will cause incorrect re-calculation of averaged
coefficients. To avoid that, either re-set model weights by assigning NULL,
or use ordered weights.
Kamil Bartoń
armWeights,
bootWeights, BGWeights, cos2Weights,
jackknifeWeights and stackingWeights can be used to
produce model weights.
weights, which extracts fitting weights from model objects.
fm1 <- glm(Prop ~ dose, data = Beetle, family = binomial) fm2 <- update(fm1, . ~ . + I(dose^2)) fm3 <- update(fm1, . ~ log(dose)) fm4 <- update(fm3, . ~ . + I(log(dose)^2)) round(Weights(AICc(fm1, fm2, fm3, fm4)), 3) am <- model.avg(fm1, fm2, fm3, fm4, rank = AICc) coef(am) # Assign equal weights to all models: Weights(am) <- rep(1, 4) # assigned weights are rescaled to sum to 1 Weights(am) coef(am) # Assign dummy weights: wts <- c(2,1,4,3) Weights(am) <- wts coef(am) # Component models are now sorted according to the new weights. # The same weights assigned again produce incorrect results! Weights(am) <- wts coef(am) # wrong! # Weights(am) <- NULL # reset to original model weights Weights(am) <- wts coef(am) # correct
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.