RG#81: plotting scatter plot with means and samples (means are connected with line while all samples as scatter plot)

set.seed(1234)
Xv <- data.frame (group = rep(1:10, each = 500),
Y = c(rnorm (500, 20, 5), rnorm (500, 35, 10), rnorm (500, 45, 15),
rnorm (500, 65, 18), rnorm (500, 50,15), rnorm( 500, 30, 10),
rnorm (500, 20, 10), rnorm (500, 20, 10),
rnorm (500, 15, 5), rnorm (500, 10,5)))

 # point plot with transparency in color
 with (Xv, plot(group, Y, pch = "-", cex=1.5, col = rgb(red=0, green=0.5, blue=0.5, alpha=0.25)))

 # calculating mean
out1 <- data.frame (with (Xv,  tapply( Y, factor(group), mean)))
names(out1) <- c("meanY")
out1$grp <- rownames (out1)

# ploting mean connected with lines
points (out1$grp, out1$meanY, type = "b", col = "red", pch = 19)

# Hexbin plot may be useful in situation of large number of data points
set.seed(1234)
Xv <- data.frame (group = rep(1:10, each = 5000),
Y = round (c(rnorm (5000, 20, 5), rnorm (5000, 35, 10), rnorm (5000, 45, 15),
rnorm (5000, 65, 18), rnorm (5000, 50,15), rnorm( 5000, 30, 10),
rnorm (5000, 20, 10), rnorm (5000, 20, 10),
rnorm (5000, 15, 5), rnorm (5000, 10,5)), 0))


   require(ggplot2)
require(hexbin)
plt <- ggplot(Xv,aes(x=group,y=Y)) + stat_binhex() + scale_fill_gradientn(colours=c("yellow","red"),name = "Frequency",na.value=NA) + theme_bw()

 # calculating mean
out1 <- data.frame (with (Xv,  tapply( Y, factor(group), mean)))
names(out1) <- c("meanY")
out1$grp <- as.numeric (rownames (out1))

# ploting mean connected with lines
plt1 <- plt + geom_point (aes(grp, meanY), data = out1, pch = 19, col = "blue", cex = 3)

# connecting with line 
plt1 +  geom_line (aes(grp, meanY), data = out1, col = "green1", lwd = 1)