code.Rmd

---
title: '2022 LJSA Data visualization with R Workshop '
author: "Sergio Uribe, @sergiouribe"
output:
  pdf_document: default
  html_notebook: default
---

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

# Preparation, load the required packages

If not installed, first install the pacman package

```{r}
# install.packages("pacman") # uncomment this line, removing the first #
```

Load the required packages

```{r}
pacman::p_load(tidyverse, palmerpenguins)
```

# Load the data

```{r}
data(penguins)
```

# Basic data exploration

# Check the structure

```{r}
str(penguins)
```

Similar to structure, but printer-friendly

```{r}
glimpse(penguins)
```

# Check the first and last rows

```{r}
head(penguins)
```

```{r}
tail(penguins)
```

# View the dataset in spreadsheet format

```{r}
View(penguins)
```

# View the names of the columns

```{r}
names(penguins)
```

# Create a summary of the dataset

```{r}
summary(penguins)
```

## Access a specific column of the dataset

Using dataset**\$**column

```{r}
summary(penguins$sex)
```

```{r}
summary(penguins$bill_length_mm)
```

# My first plot

We need

1.  data

2.  some aesthetic

3.  a geom to plot the data on the aesthetic

## The data

```{r}
penguins %>%  # load the data, and with this data >
  ggplot() # create a plot
```

## The aesthetic

```{r}
penguins %>%  # load the data, and with this data >
  ggplot(aes(x = body_mass_g))  # create a plot and add an aesthetic
```

```{r}
penguins %>%  # load the data, and with this data >
  ggplot(aes(x = body_mass_g)) + # create a plot and add an aesthetic
  geom_histogram() # add the geom, that allow to represent the data inside the aesthetic with some specified geometry
```

# Make some transformations

Usually, it is necessary to make some transformations to the data before plotting them.

The general formula is

data %\>%

**some transformation** %\>%

the plot

In this case, we are going to filter a type of penguin to see its weight distribution.

```{r}
penguins %>% # this is the data
  filter(species == "Gentoo") %>% # filter only the gentoo penguin
  ggplot(aes(x = body_mass_g)) +
  geom_histogram()
```

Alternatively, we can use all the data and make different graphs for some variables, e.g. sex and species.

```{r}
penguins %>% # this is the data
  ggplot(aes(x = body_mass_g)) +
  geom_histogram() +
  facet_grid(sex ~ species)
```

```{r}
penguins %>% 
  drop_na() %>%  # to remove the NA's we can make a transformation before plotting
  ggplot(aes(x = body_mass_g)) +
  geom_histogram() +
  facet_grid(sex ~ species)
```

# Two variables

```{r}
penguins %>% 
  ggplot(aes(x = bill_length_mm, 
         y = bill_depth_mm)) +
  geom_point()
```

## Is any correlation?

```{r}
penguins %>% 
  ggplot(aes(x = bill_length_mm, 
         y = bill_depth_mm)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE)
```

## Verify whether the correlation is robust, disaggregating by species

```{r}
penguins %>% 
  ggplot(aes(x = bill_length_mm, 
         y = bill_depth_mm, 
         shape = species, 
         color = species)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE)
```

An example of Simpon's paradox

# Polish the graph

## Change the theme

```{r}
penguins %>% 
  ggplot(aes(x = bill_length_mm, 
         y = bill_depth_mm, 
         shape = species, 
         color = species)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE) +
  # polishing the graph
  theme_classic()
```

## Add labels

```{r}
penguins %>% 
  ggplot(aes(x = bill_length_mm, 
         y = bill_depth_mm, 
         shape = species, 
         color = species)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE) +
  # polishing the graph
  theme_classic() +
  labs(title = "The Title", 
       subtitle = "Some subtitle", 
       x = "Bill length (mm)", 
       y = "Bill depth (mm)", 
       shape = "Species", 
       color = "Species")
```

# Export the plot

Check <https://ggplot2.tidyverse.org/reference/ggsave.html>

```{r}
ggsave(filename = "myFirstPlot.pdf", 
       width = 12, 
       height = 10, 
       dpi = 300, 
       units = "cm")
```