practice.Rmd

---
title: "Untitled"
author: "Author"
date: "`r format(Sys.time(), '%d %B %Y')`"
output: 
  html_document: 
    code_folding: show
    df_print: paged
    theme: cosmo
    toc: true
    toc_depth: 2
    toc_float: true
---

# Purpose
This document is an example RMarkdown to be edited and rerendered as an exercise. 
It gives examples of some good practices, like using a dynamically rendered date, 
advanced HTML options, informative headers, setting a random seed, and capturing
runtime details.

<!--
In a real analysis one would explain why this document exists and answer
questions like:
What is trying to be figured out?
How will that be done (or attempted to be done)?
What are the expected outcomes prior to the analysis?
-->

# Initialization
## Libraries
<!--
It is helpful to declare all libraries needed by a document near the top.
One may also want the chunk set to `warning=FALSE`, to avoid extra noise.
-->

```{r libs, message=FALSE}
library(magrittr)
library(tidyverse)
```

## Parameters
<!--
Values that control what and how data are processed throughout a document are 
useful to keep in one place.
This is a good place to set global theme options.
It is good practice to always declare a random numnber seed, even if unneeded.
-->

```{r params}
set.seed(20240118)

theme_set(theme_bw())
```

# Data
This document will look at the `iris` dataset.

```{r view_data}
iris
```

The observations are in centimeters and appear rounded to the nearest millimeter.

# Analysis
Here are a couple exploratory data visualizations to see the variation in 
characteristics within and across species.

## Comparing Sepal Length and Width by Species

### Point plot
```{r plt_length_v_width, fig.width=5, fig.height=4}
iris %>%
  mutate(Species=str_to_title(Species)) %>%
  ggplot(aes(x=Sepal.Width, y=Sepal.Length, color=Species)) +
  geom_point()  +
  labs(x="Sepal Width [cm]", y="Sepal Length [cm]")
```

The dots appear regularly spaced and could be prone to overplotting. That is, 
if two observations have the same measurement after rounding, then only the 
last one plotted would appear in the table. One way to deal with this is to add 
*jitter* (random uniform noise) to the data.

### Jitter plot
```{r plt_length_v_width_jitter, fig.width=5, fig.height=4}
iris %>%
  mutate(Species=str_to_title(Species)) %>%
  ggplot(aes(x=Sepal.Width, y=Sepal.Length, color=Species)) +
  geom_jitter(width=0.05, height=0.05)  +
  labs(x="Sepal Width [cm]", y="Sepal Length [cm]")
```

We use a jitter of 0.05 cm, which should be reasonable given the resolution of
the data appeared to be 0.1 cm. It should also be noted that the reproducibility
of this particular figure depends on the value given to `set.seed` in the 
[**Parameters** section](#parameters).

## Comparing Sepal Width across Species

```{r plt_box_width, fig.width=3, fig.height=3}
iris %>%
  mutate(Species=reorder(str_to_title(Species), Sepal.Width, FUN=median)) %>%
  ggplot(aes(x=Species, y=Sepal.Width)) +
  geom_boxplot() +
  labs(x=NULL, y="Sepal Width [cm]")
```

We observe the median sepal width is widest in the *I. setosa* species.

# Conclusion
This document demonstrated a few typical practices in an R Markdown document.
<!--
In a real analysis, one would clearly state each conclusion and answer 
questions like:
What was found?
Was there anything unexpected?
Are there unanswered questions?
What should be worked on next?
-->

---

# Runtime Details
## Session Info
<details>
```{r session_info, echo=FALSE}
sessionInfo()
```
</details>