05_ggplot2.Rmd

---
title: "Basic ggplot2"
author:
  - Isabel Fernandez Escapa, isabel.fernandez.escapa@gmail.com
output:
  rmdformats::robobook:
    use_bookdown: true
    code_folding: show
---

```{r klippy, echo=FALSE, include=TRUE}
klippy::klippy(c('r', 'bash'))
```

```{r, message = FALSE}
library(tidyverse)
library(Hmisc) #For summary functions
library(xtable) #For printing the statistics tables
```

# Data Import and Reshaping

```{r}
rawdata <- read.csv("data_input/Example01.csv")
```

The initial data are in **wide** format, we need them in **long format**. For this transformation we can use `reshape`:
```{r, tidy = TRUE}
longdata <- reshape(rawdata, varying =list(c("Bio.1","Bio.2","Bio.3","Bio.4")), v.names = "Biofilm", timevar = "Experiment", direction = "long")
longdata$id <- factor(longdata$id, labels = c("WT_TSB","MutA_TSB","MutB_TSB","MutAB_TSB","WT_CDM","MutA_CDM","MutB_CDM","MutAB_CDM"))
```

# Basic Plots

## Scatterplots

```{r}
ggplot(longdata) 
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm)) +
  geom_point()
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_point()
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_boxplot() 
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_boxplot() +
  geom_point(position = position_dodge(0.75), alpha=0.5)
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  geom_point(alpha=0.5) +
  stat_summary(fun = "mean", geom = "point", size = 4)
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  stat_summary(fun = "mean", geom = "point", size = 4)
```

## Bar plots

```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, color = Media)) +
  stat_summary(fun = "mean", geom = "bar")
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge())
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge()) +
  stat_summary(fun.data = "median_hilow", geom = "errorbar", position=position_dodge(.9)) 
```
```{r}
ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun.data = "median_hilow", geom = "errorbar", position=position_dodge(.9), width = 0.2) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge()) 
```

fun.data = "median_hilow" (When the default conf.int=0.95 is used, the lower and upper quantiles computed are 0.025 and 0.975.) 
See also: "mean_cl_boot" / "mean_sdl"

# Styling plots

```{r}
ugly.plot <- ggplot(longdata, aes(x = Strain, y = Biofilm, fill = Media)) +
  stat_summary(fun.data = "median_hilow", geom = "errorbar", position=position_dodge(.9), width = 0.2) +
  stat_summary(fun = "mean", geom = "bar", position = position_dodge()) 
```

```{r}
ugly.plot +
  theme_classic()
```
```{r}
ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40")) 
```
```{r}
ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40")) +
  scale_y_continuous(expand = c(0, 0)) 
```
```{r}
ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40")) +
  scale_y_continuous(expand = c(0, 0)) +
  ylab("Biofilm(OD600)") + xlab("")
```

# Facets

```{r}
ugly.plot +
  theme_classic() +
  scale_fill_manual(values = c("gray80", "gray40")) +
  scale_y_continuous(expand = c(0, 0)) +
  ylab("Biofilm(OD600)") + xlab("") +
  facet_wrap(~ Media)
```

# Statistics

## One-way ANOVA

```{r}
ANOVA <- aov(Biofilm ~ id, longdata)
```

```{r, echo = FALSE}
summary(ANOVA)
```

```{r, results = 'hide', echo = FALSE}
pvalue <- summary(ANOVA)[[1]][[1,"Pr(>F)"]]
```
**RESULT:** There is a statistically significant difference between these conditions, with a p-value of `r pvalue`

## Two-way ANOVA

```{r}
ANOVA2 <- aov(Biofilm ~ Strain + Media, longdata)
```

```{r, echo = FALSE}
summary(ANOVA2)
```

```{r, results = 'hide', echo = FALSE}
pvalue1 <- summary(ANOVA2)[[1]][[1,"Pr(>F)"]]
pvalue2 <- summary(ANOVA2)[[1]][[2,"Pr(>F)"]]
```
**RESULT:** 
There is a statistically significant difference between strains, with a p-value of `r pvalue1`. 
There is a statistically significant difference between media, with a p-value of `r pvalue2`. 

## T-test Pairwise comparisons

**Pairwise comparisons with Bonferroni correction**
```{r, results = 'hide', echo = FALSE}
pairwise.t.test(longdata$Biofilm,longdata$id, p.adj = "bonf")
```

```{r results = 'asis'}
BONF <- pairwise.t.test(longdata$Biofilm,longdata$id, p.adj = "bonf")
BONF.table <- xtable(BONF$p.value, digits = 4)
print(BONF.table, type = "html")
```

**Pairwise comparisons with Tukey correction**
```{r, results = 'hide', echo = FALSE}
TukeyHSD(ANOVA, "id")
```

```{r results = 'asis'}
TUKEY <- TukeyHSD(ANOVA, "id")
TUKEY.table <- xtable(TUKEY$id, digits = 4)
print(TUKEY.table, type = "html")
```