Plotting.Rmd

---
title: "Plotting Models"
author: "C.Stewart"
date: "2023-07-13"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```


•	do put all the model diagnostics for all the models that you present in your report in a supplementary material document.
•	make plots visualizing the relationships (e.g. plots of followed individuals that show proportion that are adult female, adult male, subadult, etc...)
•	You may also consider a plot that shows the proportion of individuals that were followed that were of the logical categories (adult male, adult female, subadult....etc...), and a similar plot for those that were following.


#LEADER plots
```{r}
rm(list=ls())

Miya_theme<-theme(axis.text = element_text(size=10, color="black"),
                  axis.title=element_text(size=10, color="black"),
                  axis.line = element_line(color="black"),
                  axis.ticks = element_line(color="black"),
                  legend.position = "top",
                  panel.background = element_rect(fill = "white", colour = "white",size = 1),
                  panel.grid = element_blank(),
                  panel.border = element_rect(colour = "black", fill=NA, size=1),
                  plot.background = element_rect(fill = "white", colour = "white",size = 1))

#set working directory
setwd("C:/Users/chars/OneDrive/Documents/CGS Project/Data")

#import csv file
df <- read.csv("summary.group.composition.plotting.csv")

#convert sex and age to characters
df$sex<- as.character(df$sex)
df$age<- as.character(df$age)

# Create an ordered factor variable for 'age'
df$age_category <- factor(df$age, levels = c(1, 2, 3),labels = c("Juvenile", "Subadult", "Adult"))

df$sex <- factor(df$sex, levels = c(1, 2), labels = c("Male", "Female"))


summary(df)

library(tidyverse)
```


# Exploratory plots (descriptive): LEADER
```{r}

# followship as a function of age (violin plot)
ggplot(data = df, aes(x = age_category, y = total_followed)) + geom_violin()

# followship as a function of sex (violin plot)
ggplot(data = df, aes(x = sex, y = total_followed)) + geom_violin()

# followship as a function of number of total adults
ggplot(data = df, aes(x = Total_Adults, y = total_followed)) + geom_point()

# there are some foraging groups where total adults = 0, 
# should these be discounted?

ggplot(data = df, aes(x = Total_Adults, y = total_followed, colour = sex)) +
  geom_count()



```

# Explore relationship between total followed and total adults
```{r}

# creating a df without the 0,0 values
# so we can examine the relationship without the zero adult points
df2 <- df[df$Total_Adults != 0, ]

LeaderPlot1<-ggplot(data = df2, aes(x = Total_Adults, y = total_followed)) +
 geom_point() + stat_smooth(se = FALSE, colour = "black", method = "lm") + Miya_theme + labs(
 x = "Number of adults per group",y="")

ggplot(data = df, aes(x = Total_Adults, y = total_followed)) +
 geom_count() + stat_smooth(se = FALSE, colour = "black", method = "lm") + Miya_theme + labs(
 x = "Number of adults per group", 
 y = "Number of leader occurences")
```


Create a plot with larger dots for higher density of points - hasn't worked
```{r}

df$coordinates <- paste(df$total_followed, df$Total_Adults, sep = ",")

df$frequency <- ave(df$coordinates, df$coordinates, FUN = length)

ggplot(data = df, aes(x = Total_Adults, y = total_followed)) +
 geom_count() + stat_smooth(method = "lm")
#tried to create plot with LM line of best fit
```


# Facet plots
```{r}

ggplot(data = df, aes(x = Total_Adults, y = total_followed, colour = sex)) +
 geom_point() +
 facet_wrap(~ age_category)
#uses age as facet, but age should not be categorical?

#plot total_followed~ age and sex only 
ggplot(data = df, aes(x = age, y = total_followed, colour = sex)) +
 geom_violin() 


LeaderPlot2<-ggplot(df, aes(x = age_category, y = total_followed, fill = sex)) +
  geom_violin() +
  labs(x = "Age", y="",fill = "Sex") +
  scale_fill_manual(values = c("Male" = "blue", "Female" = "red")) +Miya_theme
LeaderPlot2
```


# More complex stuff
```{r}

plot_model(M2b)

plot_model(M2b, type = "pred", terms = c("age","sex","Total_Adults"), 
           show.values = TRUE) +
  facet_grid(. ~ group)

#works but some funny output

plot_model(M2b, type = "pred", terms = c("age", "sex", "Total_Adults"),
           show.values = TRUE) +
  facet_wrap(sex ~ sex)

library(ggplot2)

# Obtain predicted values from the model
df$predicted <- predict(M2b, re.form = ~0)

# Create the facet plot
ggplot(df, aes(x = age, y = predicted)) +
  geom_point() +
  geom_line() +
  facet_grid(sex ~ .) +
  labs(x = "Age", y = "Predicted Values")
#kind of interesting plot ?
```

#FACET code from Miya
```{r}

library('ggplot2')
require(sjPlot)
library('ggpubr')
#create predicted plots

################################
# want four plots (facet grid) separated by sex (male, female) and age (adult, subadult)

#plot raw variables

#set variables for plots
df$sex <-as.factor(df$sex)

#plot raw values
FollowedPlot  <- ggplot(data = df, aes(x = Total_Adults, y = total_followed, colour=sex)) +
  geom_point() +
  stat_smooth(method = "lm")+
  facet_grid(sex ~ age)+
  #labs(x="Age", y="Total followed")+
  #scale_fill_grey()+
  #scale_x_continuous(breaks = seq(0, 1, by = 1))+
  Miya_theme +
  theme(legend.position = c(0.2, 0.93),legend.direction = "horizontal", legend.key.size = unit(1, 'cm'),
        legend.title = element_text(size=14), legend.text = element_text(size=14))
FollowedPlot 

#can we combine with predicted values plots?
FollowedPlotPredicted <- plot_model(M2b, type = "pred", terms = c("age","sex"),
                    axis.title = c("age", "Total followed"), title = " ") 
FollowedPlotPredicted

```

#FINAL leader plot
```{r}
library(egg)
library(cowplot)

# Combine plots using cowplot's plot_grid function
combined_plot <- plot_grid(LeaderPlot1,LeaderPlot2 ,ncol = 1,align = "v", axis = "lr",rel_heights = c(1,1))


y_axis_label <- ggdraw() +
  draw_label("Number of leader occurences", angle = 90, size = 10) +
  theme_void()

final_plot <- plot_grid(
  y_axis_label,
  combined_plot,
  ncol = 2,
  rel_widths = c(0.1, 0.9)
)


# Display the combined plot
print(final_plot)

```


#FOLLOWER plots
```{r}
# follower as a function of age (violin plot)
ggplot(data = df, aes(x = age, y = total_follower)) + geom_violin()

# follower as a function of sex (violin plot)
ggplot(data = df, aes(x = sex, y = total_follower)) + geom_violin()

# follower as a function of number of total adults
ggplot(data = df, aes(x = Total_Adults, y = total_follower)) + geom_point()

# there are some foraging groups where total adults = 0, 
# should these be discounted?

ggplot(data = df, aes(x = age, y = total_follower, colour = sex)) +
 geom_violin() 

ggplot(df, aes(x = age_category, y = total_follower, fill = sex)) +
  geom_violin() +
  labs(x = "Age", y = "Total Follower", fill = "Sex") +
  scale_fill_manual(values = c("Male" = "blue", "Female" = "pink")) + Miya_theme
  

ggplot(data = df, aes(x = total_followed, y = Total_Adults, colour = sex)) +
  geom_point()
#
```

```{r}
ggplot(data = df, aes(x = Group_Size, y = total_follower)) +
 geom_point()
```

```{r}
FollowerPlot  <- ggplot(data = df, aes(x = Group_Size, y = total_followed, colour=sex)) +
  geom_point() +
  stat_smooth(method = "lm")+
  facet_grid(sex ~ age)+
  #labs(x="Age", y="Total followed")+
  #scale_fill_grey()+
  #scale_x_continuous(breaks = seq(0, 1, by = 1))+
  Miya_theme +
  theme(legend.position = c(0.2, 0.93),legend.direction = "horizontal", legend.key.size = unit(1, 'cm'),
        legend.title = element_text(size=14), legend.text = element_text(size=14))
FollowerPlot 
```




```{r}
plot(ggpredict(M2ab))
Plot <- ggarrange (M2ab$age, M2ab$sex, ncol=2, nrow=1, common.legend = TRUE, legend = "bottom")
view(Plot)

# make one plot through using a facet command as a layer to your ggplots

 + facet_wrap(~ grouping variable)

 
# facet_grid(groupvariable1 ~ groupvariable1)£ which would be sex¬satiation level in the Zuur data.


```