effect: Plot effect for one variable from your linear regression...

In loganstundal/POL3085: Automates linear regression model plotting

Description

This function takes the linear regression model you estimate with lm() for your final projects and returns a plot of an effect for a selected variable.

Usage

effect(model, var, level = 0.95, categoricalIV = F, x.label = NULL,
  y.label = NULL, title = NULL, grid = F, rug = F,
  ci.color = "lightgray", line.width = 1, line.color = "black",
  pt.color = "black", pt.size = 1, pt.type = 16, plot.title = NULL,
  x.value.labels = NULL, connect = FALSE, connect.lwd = 1,
  connect.col = "black")

Arguments

model	The object name of your linear regression model. Example - if you run the following command: model1 <- lm(y~x), then enter 'model1' for this argument.
var	The variable to plot change in predicted values of 'y.' Your 'x' variable.
level	Confidence level, .95 is the default.
categoricalIV	Indicator for whether the plotted IV is a dummy variable, default is FALSE
x.label	Option to change x-axis label in plot
y.label	Option to change y-axis label in plot
grid	Option to plot with grid lines, default is FALSE
rug	Option to plot with rug to represent data density in plot, default is FALSE
ci.color	Color for confidence interval polygons, default is 'lightgray'.
line.width	Option to change fit regression line thickness, default is 1.
pt.color	Point color for average estimated effect of dummy independent variable, default is 'black'.
pt.size	Point size for average estimated effect of dummy independent variable, default is 1.
pt.type	Point style of regression parameter point estimate. Default is 16, circle. See ?points for information on different styles.
plot.title	Option to add title to plot
x.value.labels	Optional list of character names (must match length of categories)
connect	For categorical IV plots, add connecting line between points.
connect.lwd	Optional line width argument for connecting line in categorical IV plots.
connect.col	Optional line color option for connecting line in categorical IV plots.
plot.title	Optional plot title.

Examples

# Estimate a regression using 'mtcars' data:
m <- lm(mpg~disp+hp+wt+am,data=mtcars)

# Plot effect of 'wt' variable
effect(m,'wt')

# Plot effect of dummy variable, 'am' with custom labels
effect(m,'am',categoricalIV=TRUE,x.value.labels=c('Automatic','Manual'))

# Plot with user-defined x-axis label
effect(m,'wt',x.label='Vehicle Weight')

# Plot with grid lines and rug
effect(m,'wt',x.label='Vehicle Weight',grid=TRUE,rug=TRUE)

loganstundal/POL3085 documentation built on May 21, 2019, 10:32 a.m.