nrsq: Calculate a coefficient of determination (R-squared)

Description Usage Arguments Value See Also Examples

View source: R/nrsq.R

Description

nrsq() calculates the coefficient of determination, or R-squared value, between an independent variable x and a dependent variable y. Note that when na.rm = FALSE (default), missing values of x will be replaced with 1 and a flag identifying the missings will be included as an additive term in the model. Otherwise, only non-missing records are fit.

Usage

1
nrsq(x, y, na.rm = FALSE)

Arguments

x

numeric vector (independent variable)

y

numeric vector (dependent variable)

na.rm

logical value indicating whether missing values of x (and their corresponding y values) should be removed

Value

A numeric value with class "mt_nrsq" that indicates the proportion of the variance in the dependent variable that is predictable from the independent variable.

See Also

lm, summary.lm

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
# basic example without missing values
nrsq(mtcars$hp, mtcars$mpg)

# add some missing values
x <- ifelse(runif(length(mtcars$hp)) < 0.30, NA, mtcars$hp)
# include missings with adjustment
nrsq(x, mtcars$mpg, na.rm = FALSE)
# exclude missings entirely
nrsq(x, mtcars$mpg, na.rm = TRUE)

# evaluate a whole data frame of predictors
temp <- lapply(mtcars[c("disp", "hp", "wt", "qsec")], nrsq, y = mtcars$mpg)
data.frame(
    VarName = names(temp),
    RSquared = as.numeric(unlist(temp)),
    stringsAsFactors = FALSE
)

dnegrey/miscTools documentation built on Sept. 19, 2017, 12:15 a.m.