################################# ## ## ## CHAPTER 5: MULTIPLE SAMPLES ## ## ## ################################# ########################################################################### ## AN EXAMPLE OF ## ## MULTIPLE-SAMPLES ## ## DATA ## piglets <- read.table(file.choose(), header = T, sep = ",") # load the data in from the Excel CSV. file # 'pig_litters' and name it 'piglets' View(piglets) # view the data cleanpiglets <- lapply(piglets, function(col)col[!is.na(col)]) # remove NA values View(cleanpiglets) # view the cleaned data ########################################################################### ## ## ## BOXPLOTS ## ## ## ### Simple boxplot boxplot(cleanpiglets, # plot the data as a boxplot ylab = "Birth weight (lb)", # set y-axis label ylim=c(0,5), yaxs = "i", # specify y-axis limits col="plum") # set colour ### Refined boxplot lit1<-cleanpiglets$litter1 # give each sample to plot a convenient name lit2<-cleanpiglets$litter2 lit3<-cleanpiglets$litter3 lit4<-cleanpiglets$litter4 lit5<-cleanpiglets$litter5 lit6<-cleanpiglets$litter6 lit7<-cleanpiglets$litter7 n <- c(length(lit1),length(lit2),length(lit3),length(lit4), length(lit5),length(lit6),length(lit7)) # create list of sample sizes n # view n text(x= c(1,2,3,4,5,6,7), # add text at x-coordinates 1-7 y= 0.5, # set y-coordinate labels= c(n), # use the list of sample sizes as the text col= "purple4") # set text colour switcheddata <- cleanpiglets[c(2,6,7,3,4,5,1)] # switch the order of the samples in the data, # ordering by increasing sample size View(switcheddata) # view the switched data switchedn<- n[c(2,6,7,3,4,5,1)] # switch the order of the sample sizes list boxplot(switcheddata, # plot the switched order data as a boxplot ylab = "Birth weight (lb)", # set y-axis label ylim=c(0,5), yaxs = "i", # specify y-axis limits col="plum") # set colour ## Add the sample sizes in the new order: text(x= c(1,2,3,4,5,6,7), # add text at x-coordinates 1-7 y= 0.5, # set y-coordinate labels= c(switchedn), # use the list of switched sample sizes as the text col= "purple4") # set text colour boxplot(switcheddata, ylab = "Birth weight (lb)", ylim=c(0,5), yaxs = "i", names=c("Litter 2", "Litter 6", "Litter 7", "Litter 3", "Litter 4", "Litter 5", "Litter 1"), # re-name x-axis tick labels col="plum") text(x= c(1,2,3,4,5,6,7), y= 0.5, labels= c(switchedn), col= "purple4") ########################################################################### ## ## ## HISTOGRAMS ## ## ## ### Multiple histogram plots par(mfrow=c(3,1)) # arrange 3 plots in 3 rows with 1 column hist(lit1, # plot the data for litter 1 as a histogram col="darkgoldenrod1", # set colour xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", # set axis limits xlab="Birth weight (lb)", # set x-axis label main="Litter 1") # include the sample as a title hist(lit4, col="darkgoldenrod1",xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="Birth weight (lb)",main="Litter 4") # the same for litter 4 hist(lit7, col="darkgoldenrod1",xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="Birth weight (lb)",main="Litter 7") # the same for litter 7 par(mfrow=c(1,3)) # arrange 3 plots in 1 row with 3 columns hist(lit1, col="darkgoldenrod1",xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="Birth weight (lb)",main="Litter 1") hist(lit4, col="darkgoldenrod1",xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="Birth weight (lb)",main="Litter 4") hist(lit7, col="darkgoldenrod1",xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="Birth weight (lb)",main="Litter 7") ### Overlaying histograms in a single plot par(mfrow=c(1,1)) # return to plotting in 1 row, 1 column hist(lit1, # plot this sample as a histogram breaks=c(seq(1,5,0.5)), # specify bins col="yellow2", # set a colour for this sample xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", # set axis limits xlab="Birth weight (lb)", # set x-axis label main=NULL) # do not include a title par(new=TRUE) # prepare to draw on top of existing plot hist(lit4, # plot this sample as a histogram breaks=c(seq(1,5,0.5)), # specify bins col="springgreen3", # set a colour for this sample xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", # set axis limits xlab="",ylab="", # keep axis labels blank main=NULL) # do not include a title par(new=TRUE) # prepare to draw on top of existing plot hist(lit7, # plot this sample as a histogram breaks=c(seq(1,5,0.5)), # specify bins col="royalblue3", # set a colour for this sample xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", # set axis limits xlab="",ylab="", # keep axis labels blank main=NULL) # do not include a title legend("topright", # add a legend in the top right of the plot title="Litter", # set legend title legend=c("1","4","7"), # define sample names fill=c("yellow2","springgreen3","royalblue3"), # set legend colours in the # order of the sample names bty="n") # do not encase legend in a box install.packages("yarrr") # install package to edit transparency of colours library(yarrr) # activate package hist(lit1, breaks=c(seq(1,5,0.5)), col=transparent(orig.col = "yellow2", trans.val = 0.4), # set colour to 40% transparent yellow2 xlim=c(1,5),xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="Birth weight (lb)", main=NULL) par(new=TRUE) hist(lit4, breaks=c(seq(1,5,0.5)), col=transparent(orig.col = "springgreen3", trans.val=0.6), # set colour to 60% transparent springgreen3 xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="",ylab="",main=NULL) par(new=TRUE) hist(lit7, breaks=c(seq(1,5,0.5)), col=transparent(orig.col = "royalblue3", trans.val=0.8), # set colour to 80% transparent royalblue3 xlim=c(1,5), xaxs = "i", ylim=c(0,4), yaxs = "i", xlab="",ylab="",main=NULL) legend("topright",title="Litter", c("1","4","7"), fill=c(transparent(orig.col = "yellow2", trans.val = 0.4), transparent(orig.col = "springgreen3", trans.val=0.6), transparent(orig.col = "royalblue3", trans.val=0.8)), # set legend colours in the # order of the sample names # including transparency bty="n") ########################################################################### ## GROUPED ## ## MULTIPLE-SAMPLES ## ## DATA ## ### An example of grouped multiple-samples data # Load in the data set (seed_data.csv): seeds <- read.table(file.choose(), header = T, sep = ",") # load the data in from the Excel CSV. file # 'seed_data' and name it 'seeds' View(seeds) # view the data uncovered <- subset(seeds, treatment=="uncovered") # subset 'uncovered' data covered <- subset(seeds, treatment=="covered") # subset 'covered' data boxplot(uncovered$germinated~uncovered$water, # plot the 'uncovered' data ylab = "Seeds germinating per box", xlab = "Amount of water", # set axis labels ylim=c(0,100),yaxs = "i",xlim=c(0.7,5.3), xaxs= "i", # set axis limits xaxt="n", # do not draw x-axis col="gray", # set colour for 'uncovered' data boxwex=0.27, # make boxes narrower with scaling factor at = 1:5 - 0.15) # set where to position boxes along x-axis (slightly to the left) par(new=TRUE) # prepare to draw on top of existing plot boxplot(covered$germinated~covered$water, # plot the 'covered' data ylab = "", xlab = "", # keep axis labels blank ylim=c(0,100), yaxs = "i", xlim=c(0.7,5.3),xaxs= "i", # set axis limits yaxt="n",xaxt="n", # do not draw axes col="yellow", # set colour for the 'covered' data boxwex=0.27, # make boxes narrower with scaling factor at = 1:5 + 0.15) # set where to position boxes along x-axis (slightly to the right) axis(side=1, # draw an x-axis at the bottom of the plot at=1:5, # specify spacing of tick marks labels=c(1:5)) # label tick marks with numbers 1-5 legend("topleft", # add legend to top left of plot bty="n", # do not encase legend in box legend=c("Uncovered","Covered"), # define sample names fill=c("gray","yellow")) # set legend colours in the # order of the sample names