hedonic.function: Constructor for a "hedonic.function" Object
In hepi: Functions for estimating hedonic elementary price indices

Description Usage Arguments Details Value Author(s) See Also Examples

View source: R/hepi.R

This function constructs objects of class "hedonic.function".

1 2	hedonic.function(hf, characteristics.names, env = NULL, call = match.call(), description = "")

`hf`	The hedonic function.
`characteristics.names`	A vector of characteristics names.
`env`	An optional list of objects needed internally for the evaluation of `hf`.
`call`	An object of class `"call"`, describing how `hf` was created.
`description`	A textual description of the hedonic function `hf`.

A hedonic function predicts a price value for an item based on a list of characteristics expressions. hf is thus an object of class "function" having at least the named argument data representing the data frame for which new prices are to be estimated.

characteristics.names should contain the names of all variables that must be available in data in order to provide valid price predictions. It can be used to check whether the returned hedonic function is applicable for a given data set.

An object of class "hedonic.function".

Michael Beer r-hepi@michael.beer.name

hedonic.function-class

## Build hedonic function from training data set
build.hf.loglin <- function(traindata) {
    model <- lm(log(price) ~ ., traindata)

    hf <- function(data) {
        predict(model, newdata = data)
    }

    hedonic.function(
        hf = hf,
        characteristics.names = all.vars(formula(model)),
        env = list(model = model),
        call = match.call(),
        description = "Semi-logarithmic hedonic function."
    )
}