Regression

Francis Galton

SunflowerPlot

library(HistData)
library(car)

## Loading required package: carData

with(Galton, 
     {
       sunflowerplot(parent, child, xlim = c(63, 74), ylim = c(61, 74))
       reg <- lm(child ~ parent)
       abline(reg)
       lines(lowess(parent, child), col = "green", lwd = 2)
       if(require(car)) {
         dataEllipse(parent, child, xlim = c(63, 74), ylim = c(61, 74), plot.points = FALSE)
       }
     })

Pairs.Panels

library(psych)

## 
## Attaching package: 'psych'

## The following object is masked from 'package:car':
## 
##     logit

data(galton)
describe(galton)

##        vars   n  mean   sd median trimmed  mad  min  max range  skew
## parent    1 928 68.31 1.79   68.5   68.32 1.48 64.0 73.0     9 -0.04
## child     2 928 68.09 2.52   68.2   68.12 2.97 61.7 73.7    12 -0.09
##        kurtosis   se
## parent     0.05 0.06
## child     -0.35 0.08

#show the scatter plot and the lowess fit 
pairs.panels(galton, 
             main = "Galton's Parent child heights")  

#but this makes the regression lines look the same
pairs.panels(galton, 
             lm = TRUE, 
             xlim = c(60, 75),
             ylim = c(60, 75),
             main = "Galton's Parent child heights") 

#better is to scale them 
pairs.panels(galton,
             smooth = TRUE,
             xlim = c(60, 75),
             ylim = c(60, 75),
             main = "Galton's Parent child heights")

Pearson and Lee

dataEllipse

with(PearsonLee, 
     {
       lim <- c(52, 80)
       xv <- seq(52, 80, .5)
       sunflowerplot(parent,child, number = frequency, xlim = lim, ylim = lim, seg.col = "gray", size = .1)
       abline(lm(child ~ parent, weights = frequency), col = "blue", lwd = 2)
       lines(xv, predict(loess(child ~ parent, weights = frequency), data.frame(parent = xv)), 
             col = "red", lwd = 2)
       # NB: dataEllipse doesn't take frequency into account
       if(require(car)) {
         dataEllipse(parent, child, xlim = lim, ylim = lim, plot.points = FALSE)
       }
     })

ggplot

library(ggplot2)

## Registered S3 methods overwritten by 'ggplot2':
##   method         from 
##   [.quosures     rlang
##   c.quosures     rlang
##   print.quosures rlang

## 
## Attaching package: 'ggplot2'

## The following objects are masked from 'package:psych':
## 
##     %+%, alpha

ggplot(PearsonLee, aes(x = parent, y = child, weight = frequency)) +
  geom_point(size = 1.5, position = position_jitter(width = 0.2)) +
  geom_smooth(method = lm, 
              aes(weight = frequency, colour = "Linear"), 
              se = FALSE, size = 1.5) +
  geom_smooth(method = loess, 
              aes(weight = frequency, colour = "Loess"), 
              se = FALSE, size = 1.5) +
  facet_grid(chl ~ par) +
  scale_colour_manual(breaks = c("Linear", "Loess"),
                      values = c("green", "red")) +
  theme(legend.position = c(0.14, 0.885),
        legend.background = element_rect(fill = 'white'))