freqweights (version 0.0.1)

biglmfreq: Estimates the coefficients of a linear model

Description

Estimates the coefficients of a linear model following the guidelines of biglm

Usage

biglmfreq(formula, data, freq = ~freq)
## S3 method for class 'biglmfreq':
coef(object, ...)
## S3 method for class 'biglmfreq':
predict(object, ...)
## S3 method for class 'biglmfreq':
print(x, ...)
## S3 method for class 'biglmfreq':
update(object, ...)

Arguments

formula

a model formula

data

data frame that must contain all variables in formula and freq

freq

a one-sided, single term formula specifying frequency weights

object

a biglmfreq object

...

See Details

x

a biglmfreq object

Value

• A biglmfreq object.

Details

Any variables in the formula are removed from the data set.

It only computes the coefficients of the linear model.

... should be a data frame when predict. See Examples

... should be a data frame when update. See Examples

See Also

biglm, make.readchunk

Examples

ml <- lm(Sepal.Length ~ Sepal.Width, iris)
tt <- tablefreq(iris[,c("Sepal.Length","Sepal.Width")])
mt <- biglmfreq(Sepal.Length ~ Sepal.Width, tt)
all.equal(coef(mt), coef(ml))
coef(mt)

chunk1 <- tt[1:30,]
mf1 <- biglmfreq(Sepal.Length ~ Sepal.Width, chunk1)

chunk2 <- tt[-c(1:30),]
mf2 <- update(mf1, chunk2)

all.equal(coef(mf2), coef(ml))

predict(mf2, iris)