---
title: "Filtering, Boxplots & Histograms"
subtitle: "visualizing subsets of data"
format: html
date: today
toc: true
toc-title: Outline
---

## Warm-up

1.  What's wrong with the following code

```{r, warning=FALSE, message=FALSE}
library(tidyverse)
library(gapminder)
ggplot(gapminder,aes(x="year",y="pop")) + geom_point()
```

2.  Create an R-script that does the following.

<!-- -->

i.  Load the gapminder data
ii. Create a variable where each value is a country in the gapminder data set, with *no repeats*.
iii. Use `%in%` to see if the variable contains various countries of your choice.
iv. Note how to execute the script using the keypad, which is just one way to execute a script.

```{r}
#| eval: FALSE
source("class_9-2.R",echo = TRUE)
```

3.  Repeat the exercise above but instead use a Quarto Document.

## The filter command

Below we make a subset of the data, whose `country` is China

```{r}
C <- filter(gapminder, 
       country == "China")
```

Do a `?filter` to learn how else to modify the 2nd parameter using `& , |` and more.

Use a filter to reduce the size of the data and then label points on a scatterplot using `geom_text_repel`

```{r}
# pak::pkg_install("ggrepel")
library(ggrepel)
hi_pop_countries <- filter(gapminder, 
                           pop > 500000000)
ggplot(hi_pop_countries, 
       aes(x = year, y = gdpPercap)) + 
       geom_point() + 
       geom_text_repel(aes(label = country))
```

## boxplot & reorder

A boxplot is a visualization of the the distribution of a dataset via its *five-number summary*. ![boxplot](box-plot.png)

Here's another filter and preview of boxplots & the `reorder` function.

```{r}
hi_pop_countries <- filter(gapminder, 
                           pop > 50000000)

ggplot(hi_pop_countries, 
       aes(x = lifeExp, y = country)) + geom_boxplot()
```

Isn't this better? Use a plot & reorder for different variable.

```{r}
ggplot(hi_pop_countries, 
       aes(x = lifeExp, y = reorder(country,lifeExp))) + 
    geom_boxplot()
```

## histogram 

A histogram is another way to visualize the distribution of a dataset.  A *bin* or range of values is chosen, visible as the width of the bars.  The length of the (contiguous) bars reflects the frequency of the data within each bin.  

```{r}
TCU <- filter(gapminder,country %in% c("Taiwan","China","United States"))

TCU |> ggplot(aes(x = lifeExp, fill = country)) + geom_histogram(bins = 10)
```

## Explore the storms dataset

The `data()` command lists all dataset included with R and the Tidyverse. Note that the *storms* data is in `dplyr`.

Use filter and varioius geoms `geom_point(), geom_histogram(), geom_boxplot()`. to compare storms across time.

## Assignment 3

Complete these exercises. Use a quarto doc to generate an .html file. Copy the questions into your .qmd file and insert your responses after each one. Submit only the .html file.

Due: Midnight Tuesday 9/8

1.  Text: <https://r4ds.hadley.nz/data-visualize.html#exercises>

2.  Text: <https://r4ds.hadley.nz/data-visualize#exercises-1>

3.  Create three plots using filter and varioius geoms `geom_point(), geom_histogram(), and geom_boxplot()` to compare storms across time.