This chapter covers the following topics:
Creating R Markdown reports
Learning the markdown syntax
Embedding R code chunks
Creating interactive graphics with ggvis
Understanding basic syntax and grammar
Controlling axes and legends
Using scales
Adding interactivity to a ggvis plot
Creating an R Shiny document
Publishing an R Shiny report
After completing data analysis, it is important to document the research results and share the findings with others. The most common methods involve documenting results through text, slides, or web pages. However, these formats are generally limited to only sharing the results and don't often include the research process. As a result, other researchers cannot reproduce the research, making it hard to fully understand how the author conducted the analysis. Without knowing how to replicate the research, the authenticity of the study may be questioned.
R scripts can be used to perform data analysis and generate figures, and it is possible to create a reproducible report by copying all the code and images into a document. However, as this is quite labor intensive, there is a risk of making errors. A better solution is to automate the documentation process. This allows the user to dynamically generate a report, in any format, which records both scripts and the analysis results.
There...
RStudio has an R Markdown workflow built in; we can use its GUI to create markdown reports in HTML, PDF, slide, or Microsoft Word format. In this recipe, we will introduce how to build an R Markdown report with RStudio.
Ensure you have installed the latest versions of R and RStudio on your operating system. If you have not yet installed RStudio, please visit the following URL to download the most recent version:
Please perform the following steps to create an R Markdown report:
First, click on File | New File | R Markdown on the menu bar:
Figure 1: Create an R Markdown report
Next, select the document type in the left-hand side menu, then fill in the Title and Author, and choose the output format in the right-hand side menu:
Figure 2: Fill in the meta information of an R Markdown report
Moving on, we can now see a markdown document with a YAML header and content in markdown syntax. You can now...
Markdown is a lightweight markup language; it allows anyone to quickly create a formatted document in plain text instead of using complicated HTML. In this recipe, we introduce how to write a report with markdown syntax.
Ensure you have installed the latest version of R and RStudio on your operating system. Also, you need to have created and opened a new R Markdown (.rmd) file in RStudio.
Please perform the following steps to write simple markdown syntax:
First, create headings by placing a hashtag # in front of the words:
|
Markdown |
Preview |
|---|---|
|
|
 |
Next, emphasize words by style with bold, italics, and strikethrough:
|
Markdown |
Preview |
|---|---|
|
This is plan text:
|
 |
For list items, we can list items in an ordered item list or an unordered item list:
|
Markdown |
Preview |
|---|---|
|
|
Markdown |
Preview |
|---|---|
|
|
 |
We can hide the script by setting echo=FALSE:
|
Markdown |
Preview |
|---|---|
|
|
 |
Alternatively, we can stop evaluating the code by setting eval=FALSE:
|
Markdown |
Preview |
|---|---|
|
|
 |
Moreover, we can choose not to render both evaluation result and script by setting include...
In order to interact with the reports figures, one can create an interactive graphic with ggvis. In this recipe, we demonstrate how to build our first interactive plot from the real estate dataset.
Before starting this recipe, you should download the RealEstate.csv dataset from the following GitHub link:
https://github.com/ywchiu/rcookbook/blob/master/chapter8/RealEstate.csv
Please perform the following steps to create an interactive plot with ggvis:
Install and load the ggvis package:
> install.packages("ggvis") > library(ggvis)
Import RealEstate.csv into an R session:
> house <- read.csv('RealEstate.csv', header=TRUE) > str(house) 'data.frame': 781 obs. of 8 variables: $ MLS : int 132842 134364 135141 135712 136282 136431 137036 137090 137159 137570 ... $ Location : Factor w/ 54 levels " Arroyo Grande",..: 21 44 44 39 50 42 50 50 39 22 ... $ Price : num 795000 399000 545000 909000 109900 ...
The ggvis uses similar grammar and syntax to ggplot2, and we can use this basic syntax to create figures. In this recipe, we cover how to use ggvis syntax and grammar to build advanced plots.
Ensure you have installed and loaded ggvis into your R session. Also, you need to complete the previous steps by storing house in your R environment.
Please perform the following steps to create plots with ggvis:
First, create a scatterplot by mapping Size to the x axis and Price to the y axis. Furthermore, we can assign different colors or shapes to points with different statuses:
> house %>% ggvis(~Size, ~Price, fill=~Status, size=10, shape=~Status) %>% layer_points()
Figure 11: A scatterplot in different shapes and colors by housing status
Also, we can use the add_props function to change the fill color to red:
> house %>% ggvis(~Size, ~Price, fill=~Status, size=10, shape=~Status) %>% layer_points() %>% add_props(fill...Besides making different plots with various layer types, one can control the axes and legends of a ggvis plot. In this recipe, we demonstrate how to set the appearance properties of both axes and legends.
Ensure you have installed and loaded ggvis into your R session. You also need to complete the previous steps by storing house in your R environment.
Please perform the following steps to control axes and legends in ggvis:
First, we use the add_axis function to control the axis orientation and label of a ggvis plot:
> house %>% ggvis(~Size, ~Price) %>% layer_points() %>% + add_axis("x", title = "Real Estate Square Feet", orient="top") %>% + add_axis("y", title = "Real Estate Price", title_offset = 80)
Figure 20: Creating an axis on the ggvis plot
We can control the scale of the axis:
> house %>% ggvis(~Size, ~Price) %>% layer_points() %>% + add_axis("x", title = "Real Estate Square Feet", + ...
Besides changing the appearance of axes and legends, we can rescale the mapping of the data and how it should be displayed on the plot with the scale function. In this recipe, we introduce how to scale data in ggvis.
Ensure you have installed and loaded ggvis into your R session. Also, you need to complete the previous steps by storing house in your R environment.
Please perform the following steps to rescale data in ggvis:
First, create a bar plot and then replace the linear scale with a power scale:
> house %>% ggvis(~Status, ~Price, fill=~Status) %>% layer_bars() %>% scale_numeric("y", trans = "pow", exponent = 0.2)
Figure 26. Adjusting the scale of the bar plot
We can change the fill color with the scale_nominal function:
> house %>% ggvis(~Status, ~Price, fill=~Status) %>% layer_bars() %>% scale_nominal("fill", range = c("pink", "green", "lightblue"))
Figure 27. Changing the fill color of the bar plot
One of the most attractive features of ggvis is that it can be used to create an interactive web form. This allows the user to subset the data, or even change the visual properties of the plot, through interacting with the web form. In this recipe, we introduce how to add interactivity to a ggvis object.
Ensure you have installed and loaded ggvis into your R session. Also, you need to complete the previous steps by storing house in your R environment.
Please perform the following steps to add interactivity to ggvis:
First, make a bar plot with a drop-down menu and color options:
> house %>% + ggvis(~Status, ~Price, fill:= input_select(c("red","blue"), label="Fill Color")) %>% + layer_bars() %>% + add_axis("x", title = "Status", title_offset = 40) %>% + add_axis("y", title = "Price", title_offset = 80)
Figure 28: Adding a drop-down menu with color options
Having introduced how to make an R Markdown document and ggvis plot, we can now combine these two together in a single report. However, the interactivity property of ggvis does not work with a regular R Markdown report. Instead, we can only enable the interactivity of the plot in a Shiny document. In this recipe, we introduce how to create a Shiny document.
Ensure you have installed and loaded ggvis into your R session. Also, you need to complete the previous steps by storing house in your R environment.
Please perform the following steps to create a Shiny document:
First, click File | New File | R Markdown on the menu bar:
Figure 35: Create an R Markdown file
Next, choose Shiny as the document type in the left-hand side menu, and then fill in the Title and Author, and select the output format in the right-hand side menu:
Figure 36: Input title and author of the R Markdown report
Type the code that generates a ggvis plot into R chunks:
```...After creating our interactive report with the Shiny application, we can publish our report online. In this recipe, we introduce how to publish an R Shiny report online.
Ensure you have installed and loaded ggvis into your R session. Also, you need to complete the previous steps by storing house in your R environment.
Please perform the following steps to publish your Shiny document online:
First, click Publish Document on the upper-right side of the document:
Figure 40: Publish the Shiny document
If you have not installed the required packages, you should click Yes to install the required packages for publishing the Shiny document:
Figure 41: Install the required packages to publish Shiny app
Click on Publish just this document:
Figure 42: Click on publish just this document
At this point, RStudio requires you to connect your ShinyApps account with app tokens. Therefore, you should log in to your account to retrieve the token:
Figure 43: ShinyApps...
© 2016 Packt Publishing Limited All Rights Reserved
