MuMIn (version 1.14.0)

predict_avg: Average predictions (experimental)

Description

Average predictions (experimental functions under development)

Usage

predict_avg(x, newdata, type = c("response", "invlink", "link", "terms"), 
    se.fit = FALSE, full = TRUE, linkinv = NULL, use.lincomb = FALSE, ...)
	
avg.predictions(yall, w, type = c("atomic", "matrix", "se.fit", "terms"), 
    revised.var = TRUE, full = FALSE, ...)## S3 method for class 'lm':
std_predict(object, newdata, type, se.fit, ...)

Arguments

x,object

model averaging object.

yall

list of predictions.

newdata

new data.

type

type of prediction.

se.fit

return SE?

w

weights

revised.var

revised.var?

full

logical, full or subset prediction

linkinv

inverse link (as function, family, link-glm, or character). Caution! linkinv = "log" will transform the prediction on a link scale using exponential function (via make.link("log"), whereas linkinv = log will

use.lincomb

logical, whether to predict using linear combination of model-averaged coefficients? Note that this currently does not compute predictions properly for models using fitting weights and offset. Component model must have model.matrix

...

other arguments...

Value

• TODO

encoding

utf-8

Details

TODO

See Also

predict.averaging, predict, predict.glm

Examples

# TODO