12.1 Reports

The main package to generate reports is knitr which - among a lot of other useful features - handles the transition from our R Markdown (.Rmd) files to Markdown files (.md). These files and its correspoding markdown language are widely used, e.g, on GitHub and stackoverflow, and allow formatting code and text via a simple syntax.

Once we have a markdown file, it is possible to convert it into a myriad of formats via pandoc. In the end, we do not really need to worry about how these languages interplay. In RStudio, we can simply hit the Knit button to start the conversion. But what we need to know is how to use Markdown’s syntax to format our texts.

# headline    level 1 header - chapter
## headline   level 2 header - section
### headline  level 3 header - subsection

* List entry 1

* List entry 2

  * Two spaces for indentation of bullets

1. List entry 1

1. List entry 2 (numbering works automatically)

![optional caption text](path/to/img.png)

Col1  | Col2
----- | -----
1     | 2
3     | 4

knitr::include_graphics("Images/taylor-tree.png")

tbl <- tibble::tribble(
  ~col1, ~col2,
  1, 2,
  3, 4
) 

tbl %>% knitr::kable()

library(kableExtra)
tbl %>% 
  knitr::kable() %>% 
  kable_classic(full_width = F)

tbl %>% 
  knitr::kable() %>% 
  kable_minimal(full_width = F)

tbl %>% 
  knitr::kable() %>% 
  kable_styling(
    # See documentation for more options
    full_width = F, 
    position = "left",
    row_label_position = "r"
  )

tbl %>% 
  knitr::kable() %>% 
  kable_styling(
    # See documentation for more options
    full_width = F, 
    position = "left",
    row_label_position = "r"
  ) %>% 
  column_spec(column = 2, bold = T) 

```{r}
```

```{r unnamed-chunk-xx, 
  eval = TRUE, 
  include = TRUE, 
  echo = TRUE, 
  message = TRUE, 
  warning = TRUE}
```

knitr::opts_chunk$set(<option> = <value>)

---
title: "R Notebook"
output: html_notebook
---

---
title: "R Notebook"
output:
  prettydoc::html_pretty:
    theme: cayman
  html_document: default
  pdf_document: default
---

---
title: "R Notebook"
author: John Jacob Jingleheimer Schmidt
date: "2021-06-16"
output:
  prettydoc::html_pretty:
    theme: cayman
    toc: true
  html_document: 
    toc: true
  pdf_document: default
---

12.2 Flexdashboards

Flexdashboards from the flexdashboard package are comparatively simple dashboards compared to the complex dashboards you can create with the shiny package (which we will work with in the next chapter). To create a flexdashboard simply install the corresponding package and change the output in the YAML header to flexdashboard::flex_dashboard.

Then, you can easily fill pages, columns and rows of this dashboard with content. All you need to know for that is that the Markdown headers are used to indicate what page (#), column (##) or row (###) you are working on. As an example, simply take the following code and knit it. Also, feel free to tweak with the code to see how your changes affect the output.

#> ---
#> title: "Flexdashboard Example"
#> output: flexdashboard::flex_dashboard
#> ---
#> 
#> ```{r}
#> library(tidyverse)
#> knitr::opts_chunk$set(fig.width = 10)
#> ```
#> 
#> # First Page (Two-columned)
#> 
#> ## Iris data 
#> 
#> ### Data
#> 
#> ```{r, }
#> iris %>% 
#>   DT::datatable() # make it an interactive table with DT package
#> ```
#> 
#> ## Iris plot
#> 
#> ### Plot1
#> 
#> ```{r}
#> ggplot(iris) +
#>   geom_point(aes(Sepal.Width, Petal.Length, col = Species))
#> ```
#> 
#> ### Plot2
#> 
#> ```{r}
#> ggplot(iris) +
#>   geom_point(aes(Sepal.Length, Petal.Length, col = Species))
#> ```
#> 
#> ### Plot3
#> 
#> ```{r}
#> ggplot(iris) +
#>   geom_point(aes(Sepal.Width, Sepal.Length, col = Species))
#> ```
#> 
#> # Second Page

Alternatively, you may use the alternative syntax that uses multiple - signs as column indicators and multiple = as page indicator. Thus, using the alternative syntax for the above dashboard the code will look like this.

#> ---
#> title: "Flexdashboard Example"
#> output: flexdashboard::flex_dashboard
#> ---
#> 
#> ```{r}
#> library(tidyverse)
#> knitr::opts_chunk$set(fig.width = 10)
#> ```
#> 
#> First Page (Two-columned)
#> =======================================
#> 
#> Iris data
#> ---------------------------------------
#> 
#> ### Data
#> 
#> ```{r, }
#> iris %>% 
#>   DT::datatable() # make it an interactive table with DT package
#> ```
#> 
#> Iris plot
#> ---------------------------------------
#> 
#> ### Plot1
#> 
#> ```{r}
#> ggplot(iris) +
#>   geom_point(aes(Sepal.Width, Petal.Length, col = Species))
#> ```
#> 
#> ### Plot2
#> 
#> ```{r}
#> ggplot(iris) +
#>   geom_point(aes(Sepal.Length, Petal.Length, col = Species))
#> ```
#> 
#> ### Plot3
#> 
#> ```{r}
#> ggplot(iris) +
#>   geom_point(aes(Sepal.Width, Sepal.Length, col = Species))
#> ```
#> 
#> Second Page 
#> =======================================

12.3 Slides

My package of choice for creating slides in R is the xaringan package which comes from the same creator as the knitr package. Apparently, the creator of this package is a huge Naruto fan which is why the package is names after the Sharingan from Naruto. Basically, the Sharingan enables one to use both the “eye of insight” and the “eye of hypnotism”. So, good luck hypnotizing your audience with your dazzling slides.

Again, using this new format is only a matter of adjusting the YAML header. Then, slides can be created via ---. Also, what is nice about these slides is that they are able to go into presentation mode (by opening the knitted html-file in your browser and pressing P) so that one can hide notes for a particular slide using ???.

An example presentation would look like this. Simply copy and paste the code into an .Rmd-file and knit it.

#> ---
#> title: "Example Slides"
#> subtitle: "With a super amazing subtitle"
#> output:
#>   xaringan::moon_reader:
#>     lib_dir: libs
#>     nature:
#>       highlightStyle: github
#>       highlightLines: true
#>       countIncrementalSlides: false
#> ---
#> 
#> ```{r, echo = F, message = F, warning = F, fig.align='center'}
#> library(tidyverse)
#> knitr::opts_chunk$set(fig.width = 10, fig.height = 5, fig.align = 'center')
#> ```
#> 
#> # Headline
#> ```{r}
#> iris %>% 
#>   ggplot(aes(Sepal.Length, Sepal.Width, col = Species)) +
#>   geom_point()
#> ```
#> 
#> ---
#> class:middle
#> 
#> ## Centering
#> This slide is now filled from the center due to `class:middle`
#> 
#> ???
#> 
#> * Using `???` I can hide notes.
#> * I can access these from my knitted html document by pressing `P` 
#> (i.e. going into presentation mode.) 
#> * Also I can clone the presentation by pressing `C`
#> 
#> ---
#> class:middle
#> 
#> ### Line highlighting via #<<
#> 
#> ```{r, eval = F}
#> mpg %>% 
#>   ggplot(aes(hwy, cty, col = class)) +
#>   geom_jitter() #<<
#> # Figure is shown on next slide! (due to --)
#> ```
#> 
#> --
#> 
#> ```{r, echo = F}
#> mpg %>% 
#>   ggplot(aes(hwy, cty, col = class)) +
#>   geom_jitter() #<<
#> ```
#> 
#> More information in the [original intro presentation](https://slides.yihui.org/xaringan/#1).

12.4 GitHub Pages

GitHub Pages is a simple but effective way to create websites from a GitHub repository without any knowledge about html, css and all else that is involved in the deployment of a website. Luckily, all that you need to know is how to set up GitHub Pages and then you can simply host your website from a GitHub repository.

Then, every time you want to change something on your website, you simply push changes into your repository and your website is built anew automatically. Also, what is great is that you can also simply add html-files that are created from knitting an Rmd-file to your repository and these automatically become accessible on your website. Thus, this is a simple way to share your reports, slides or whatever it is that you built with R.

Now, setting up a GitHub Pages site is as easy as creating a new repository. On GitHub, simply create a new public repository named <username>.github.io where <username> must correspond to your GitHub account name. Also, this repository should include a blank README file which will initialize the repository.

Once you have done that, a few seconds later, your GitHub Pages site becomes available online at <username>.github.io. If you take a look at this website, you will see that there is not much there yet. Now, you could fill your repository with all kinds of html- and css-files to bring your website to life. Alternatively, you can take the simple route by using an available theme.

In order to go with the latter option, simply go to your GitHub repository and follow Settings > Pages > Choose a theme. There, you can preview and select a theme that you fancy and then hit the Select theme button. As you will see, this also changes your README file whose changes you can commit in the next window that appears. Afterwards, you will see that your Github Pages site now looks the way it did when you were previewing themes.

Of course, now you can fill your repository with files from your R project. Here I have set up a dummy GitHub Page site that includes the dashboard and slide examples from the present chapter. As you can see from the corresponding repository I simply pushed a folder data to the repository that includes the content I want to host online.

All of that content is then accessible online via <username>.github.io/path/to/file. So, to access the flexdashboard from before, simply follow https://albertrapp.github.io/data/flexdashboard_example.html.

Of course, this gives you the opportunity to keep your README file to manage the landing page when someone follows <username>.github.io to your website. From there you can then put hyperlinks to your content in the data directory as I have done on the dummy site.

Finally, notice that a lot of knitted files are self-contained, so it is only a matter of uploading a single html-file. However, some files such as our xaringan example also include css-files for the styles of the file. These need to be uploaded as well which is why there is also a libs directory in the data directory of the dummy repository.

12.5 Exercises