#UPLOADING DATA####
#set working directory
setwd ('C:\\Users\\kabur\\Dropbox\\R workshop\\2021-22 series\\session 2')

#upload data
monkey_data <- read.csv ("treatment.csv", header = T, stringsAsFactors = F)

#let's change the name of the header from "Intervention" to "Int"
colnames (monkey_data)[colnames(monkey_data)=="Intervention"] <- "Int"

#HISTOGRAM####
#we can check the distribution of Play data using a histogram
hist (monkey_data$play)

#you can edit the histogram by determining the width of the bars (using "breaks"), 
#adding the labels (xlab), 
#determining the limits of the x or y axis (xlim, ylim), 
#adding a title (main), 
#and changing the colour (col)
hist (monkey_data$play, breaks = 20, xlab = "Play rates", 
      xlim = c(0,240), main = "Histogram of Play rates", col = "lightblue")

#BAR GRAPH####
#to plot differences in mean of a continuous variable between two groups we can use bar graphs
#let's suppose we want to compare Play rates betwen males and females
#here are the steps
#STEP 1: convert sex from number to character
monkey_data["Sex"] [monkey_data["Sex"] == 0] <- "am"
monkey_data["Sex"] [monkey_data["Sex"] == 1] <- "af"

#STEP 2: let's calculate mean and sd hccs for each sex
install.packages ('plyr')
library (plyr)
data_summary <- ddply(monkey_data, .(Sex), summarize, mean = round (mean(play),2), sd = sd(play))

#plot a basic bar graph using "ggplot"
#install ggplot2 package
install.packages ('ggplot2')
library (ggplot2)

#bar graph
p <- ggplot (data_summary, aes (x = Sex, y = mean)) + geom_bar (stat = "identity", fill = "white", colour = "black")+   
  geom_errorbar( aes(ymin=mean-sd, ymax=mean+sd),  width = .2) 

#you can add labels
q <- p +ggtitle("Difference in play rates \n between males and females")+ 
  xlab("Sex") + ylab("Play rates (N/hr)") 

#you can add text to bins
m <- q +geom_text (aes  (label= mean), vjust = 1.5, hjust = 1, size = 7)

#you can edit format of text and background as follows
w <-m + theme(plot.title = element_text(color = "black", size = 20, hjust = .5), 
          axis.text = element_text(color = "black", size = 20, hjust = .5, vjust = .5, face = "plain"), 
          axis.title= element_text(color = "black", size = 20, vjust =3),
          panel.background = element_rect(fill = "transparent",colour = NA), 
          axis.line = element_line(color="black", size = .2))

#we can also create bar graphs by grouping bars together
#let's suppose we want to plot mean play rates for intervention split by sex

#STEP 1: convert the variable "Intervention" from number to character
monkey_data$Int<- as.character (monkey_data$Int)

#STEP 2: relable each intervention into a more informative variable
monkey_data["Int"] [monkey_data["Int"] == "0"] <- "Control"
monkey_data["Int"] [monkey_data["Int"] == "1"] <- "Handling-only"
monkey_data["Int"] [monkey_data["Int"] == "2"] <- "Face-to-face"

#STEP2: calculate mean play rates by intervention and sex
data_summary2 <- ddply(monkey_data, .(Sex, Int), summarize,   
                      mean = mean(play),
                      sd = sd(play))

#STEP 3: plot
p2 <- ggplot (data_summary2, aes (x = Int, y = mean, fill = Sex)) + geom_bar (position = "dodge", stat = "identity")+
  geom_errorbar(aes(ymin=mean-sd, ymax=mean+sd), width=.2,
                position=position_dodge(.9))

#add labels and change text in legend and change colour of variables
q2 <- p2 +ggtitle("Difference in play rates \n between intervention measures split by sex")+ 
  xlab("Sex") + ylab("Play rates (N/hr.)") +
  scale_fill_manual(labels = c("Adult \n females", "Adult \n males"), values = c("green", "blue")) 

#edit format text
w2 <-q2 + theme(plot.title = element_text(color = "black", size = 20, hjust = .5), 
              axis.text = element_text(color = "black", size = 20, hjust = .5, vjust = .5, face = "plain"), 
              axis.title= element_text(color = "black", size = 20, vjust =3),
              panel.background = element_rect(fill = "transparent",colour = NA), 
              axis.line = element_line(color="black", size = .2),
              legend.title = element_text(color = "black", size = 20),
              legend.text = element_text(color = "black", size = 15, margin = margin(t=6)))

#BOXPLOT#####
#bar graph
p3 <- ggplot (monkey_data, aes (x = Int, y = play)) + geom_boxplot ()

#add labels
q3 <- p3 +ggtitle("Difference in play rates \n between intervention measures")+ 
  xlab("Intervention measures") + ylab("Play rates (N/hr.)")

#edit format text
w3 <-q3 + theme(plot.title = element_text(color = "black", size = 20, hjust = .5), 
                axis.text = element_text(color = "black", size = 20, hjust = .5, vjust = .5, face = "plain"), 
                axis.title= element_text(color = "black", size = 20, vjust =3),
                panel.background = element_rect(fill = "transparent",colour = NA), 
                axis.line = element_line(color="black", size = .2),
                legend.title = element_text(color = "black", size = 20))

#we can rearrange the order in which elements are plotted
monkey_data$Int <- factor (c ("Control", "Face-to-face", "Handling-only"), levels =c ("Face-to-face", "Handling-only", "Control"))

w3b<- ggplot (monkey_data, aes (x = Int, y = play)) + geom_boxplot () +
  ggtitle("Difference in play rates \n between intervention measures")+ 
  xlab("Sex") + ylab("Play rates (N/hr.)") +
  scale_fill_discrete(labels = c("Adult \n females", "Adult \n males")) +
  theme(plot.title = element_text(color = "black", size = 20, hjust = .5), 
        axis.text = element_text(color = "black", size = 20, hjust = .5, vjust = .5, face = "plain"), 
        axis.title= element_text(color = "black", size = 20, vjust =3),
        panel.background = element_rect(fill = "transparent",colour = NA), 
        axis.line = element_line(color="black", size = .2),
        legend.title = element_text(color = "black", size = 20),
        legend.text = element_text(color = "black", size = 15, margin = margin(t=6)))

#if you ran stats to compare differences, you can add text on the graph to indicate which pairs are significantly different
t <- w3b +  annotate ("segment", x = (2), xend = (3), y = 300, yend = 300 )+
  annotate ("segment", x = (1), xend = (3), y = 350, yend = 350) +
  annotate ("segment", x = (1), xend = (2), y = 230, yend = 230) +
  annotate ("text", x = 2.5, y =330, label ="***", size = 5) + 
  annotate ("text", x = 2, y = 380, label ='***', size = 5) +
  annotate ("text", x = 1.5, y = 250, label = "ns", size = 5)

#SCATTERPLOT####
#To plot the relationship between two continuous variables you can use the scatterplot
#let's upload a new dataset "HVS"

HVS <- read.csv ("HVS.csv", header = T)

#let's plot Orbital Volume vs Cranial Capacity
p4<-ggplot(HVS, aes(CranialCapacity, MeanOrbitalVolume))+ geom_point()

#you can add the best fit line
q4<- p4 + geom_smooth(method='lm', se = F, col = "black")

#you can edit axis labels and graph title
w4 <- q4 + xlab ("Cranial Capacity") + ylab ("Orbital Volume") + ggtitle ("Relationship between Orbital Volume and Cranial Capacity")

#let's change the format of text
c <- w4 + theme(plot.title = element_text(color = "black", size = 20, hjust = .5), 
      axis.text= element_text(color = "black", size = 20, hjust =1, vjust = .5, face = "plain"), 
      axis.title = element_text(color = "black", size = 20, vjust = 2, face = "plain"),
      panel.background = element_rect(fill = "transparent",colour = NA), 
      axis.line = element_line(color="black", size = .2))

#we can also generate two best fit line for each gender separately
ggplot(HVS, aes(CranialCapacity, MeanOrbitalVolume, colour = Gender))+ geom_point() + 
  geom_smooth(method='lm', se = F) +
  xlab ("Cranial Capacity") + ylab ("Orbital Volume") + ggtitle ("Relationship between Orbital Volume and Cranial Capacity")+
  theme(plot.title = element_text(color = "black", size = 20, hjust = .5), 
        axis.text= element_text(color = "black", size = 20, hjust =1, vjust = .5, face = "plain"), 
        axis.title = element_text(color = "black", size = 20, vjust = 2, face = "plain"),
        panel.background = element_rect(fill = "transparent",colour = NA), 
        axis.line = element_line(color="black", size = .2))
  
#we can having multiple plots together using "patchwork"
install.packages ("patchwork")
library (patchwork)
c + t

#using inset_element you can insert a graph within a graph
c + inset_element(t, 0.6, 0.6, 1, 1)