autoplot.partialResiduals: Display an first order interaction for categorical variables

In bozenne/butils: Various useful functions

Display an first order interaction for categorical variables

Description

Display an first order interaction for categorical variables

Usage

## S3 method for class 'partialResiduals'
autoplot(
  object,
  size.point = 2,
  col.point = NULL,
  dodge = 0.1,
  size.fit = NULL,
  shape.fit = NULL,
  col.fit = NULL,
  size.ci = 0.25,
  alpha.ggplot = 0.25,
  col.ci = NULL,
  plot = TRUE,
  ...
)

Arguments

size.point	[numeric, >0] size of the dots representing the observed data.
col.point	[character vector] color of the dots representing the observed data.
size.fit	[numeric, >0] thickness of the regression line.
shape.fit	[integer, >0] Symbol used to represent the fitted value.
col.fit	[character vector] color of the regression line.
size.ci	[numeric, >0] thickness of the line representing the confidence interval.
alpha.ggplot	[numeric, 0-1] transparency parameter for the confidence interval.
col.ci	[character vector] color of the line/band representing the confidence interval.
plot	[logical]should the plot be displayed?
...	ignored.
x	a linear model

Examples

library(lava)
set.seed(10)
m.lvm <- lvm(Y~2*Age+4*gender+gene+time)
categorical(m.lvm, labels = c("M","F")) <- ~gender
categorical(m.lvm, K = 10) <- ~gene
d <- lava::sim(n = 1e2, m.lvm)
d$gene <- as.character(d$gene)

## linear model
m <- lm(Y~Age+gender, data = d)
pres1 <- partialResiduals(m, var = "Age")
autoplot(pres1)
autoplot(pres1, col.point = "pink", col.ci = "orange", col.fit = "purple")
pres2 <- partialResiduals(m, var = c("Age","gender"))
autoplot(pres2)
pres3 <- partialResiduals(m, var = c("Age","gender"), interval = "prediction")
autoplot(pres3)
pres4 <- partialResiduals(m, var = "gender")
autoplot(pres4)
autoplot(pres4, col.point = "red", col.ci = "orange", col.fit = "purple")
m2 <- lm(Y~Age+gender+gene, data = d)
pres5 <- partialResiduals(m2, var = c("gender","gene"))
autoplot(pres5)
autoplot(pres5, dodge = 1, col.point = rep("black",10))

## linear mixed model
if(require(nlme)){
mm <- lme(Y~Age+gender, random= ~ 1|gene, data = d)
pres1 <- partialResiduals(mm, var = "Age")
autoplot(pres1)
}
if(require(lme4) && require(merTools) && require(AICcmodavg)){

mm <- lmer(Y~Age+gender+(1|gene), data = d)
pres1 <- partialResiduals(mm, var = "Age")
autoplot(pres1)
pres2 <- partialResiduals(mm, var = c("Age","gender"))
autoplot(pres2)

# using external function
pres3 <- partialResiduals(mm, var = c("Age","gender"), FUN.predict = predict_merTools)
autoplot(pres3) 
pres4 <- partialResiduals(mm, var = c("Age","gender"), FUN.predict = predict_AICcmodavg)
autoplot(pres4)
}

## gam
if(require(mgcv)){
set.seed(2) ## simulate some data
dat <- gamSim(1,n=400,dist="normal",scale=2)
b <- gam(y~s(x0)+s(x1)+s(x2)+s(x3),data=dat)
b <- gam(y~s(x0)+x1+x2+x3,data=dat)
pres5 <- partialResiduals(b, var = "x0")
autoplot(pres5)
}

bozenne/butils documentation built on Oct. 14, 2023, 6:19 a.m.