regplot: Scatter plot with regression lines

In timflutre/rutilstimflutre: Timothee Flutre's personal R code

Scatter plot with regression lines

Description

Make a scatter plot of y as a function of x, along with regression line(s).

Usage

regplot(
  x,
  y,
  reg = "lm",
  col = c(lm = "red"),
  show.crit = c("corP", "corS"),
  ci.int = TRUE,
  pred.int = TRUE,
  legend.x = "right",
  legend.y = NULL,
  las = 1,
  ...
)

Arguments

x	vector or 1-column matrix (missing data coded as NA will be automatically discarded)
y	vector or 1-column matrix (missing data coded as NA will be automatically discarded)
reg	specifies which model(s) to use to add regression lines(s) (lm/loess/rlm; can be c("lm", "rlm"))
col	named vector specifying the color(s) of the regression line(s) specified via reg
show.crit	specifies which criterion to show in the top left corner (corP/corS/R2adj); can be c("R2adj", "corP")) or NULL
ci.int	if reg="lm", add lines corresponding to confidence intervals
pred.int	if reg="lm", add lines corresponding to prediction intervals
legend.x	the x coordinate to be used to position the legend (see legend)
legend.y	the y coordinate to be used to position the legend (see legend)
las	see par
...	arguments to be passed to plot

Value

list of object(s) returned by the function(s) specified via reg

Author(s)

Timothee Flutre

Examples

set.seed(1859)
n <- 500
x <- rnorm(n=n, mean=37, sd=3)
y <- 50 + 1.2 * x + rnorm(n=n, mean=0, sd=3)
fit <- regplot(x=x, y=y, reg="lm", las=1, main="Linear regression")
fit <- regplot(x=x, y=y, reg="loess", las=1, col=c(loess="red"),
               main="Locally weighted scatterplot smoothing (loess)")
y2 <- y + sample(x=c(rep(0, 0.7*floor(n)),
                     rnorm(n=ceiling(0.3*n), mean=c(7,13), sd=20)), size=n)
fit <- regplot(x=x, y=y2, reg=c("lm","rlm"), las=1,
               col=c(lm="red", rlm="blue"), legend.x="bottomright",
               main="(Robust) linear regressions")

timflutre/rutilstimflutre documentation built on Feb. 7, 2024, 8:17 a.m.