In this lesson, we discuss methods for presenting research results visually. For the final project, you have the option of either creating an academic conference poster or an interactive digital dashboard.
Academic posters are designed to quickly convey research questions and key results in a visual format. The poster should be visually compelling and use minimal text.
Posters can be created with many different software programs such as InDesign, Google slides or Powerpoint. A simple way to create a poster is to use a single powerpoint slide with the document dimensions set to the dimensions that the poster will be printed in. Different conferences have different requirements for poster size. For example, the AAA maximum poster size is 4ft x 6ft. Working on such a large document, your headings and text will also be much larger.
A good rule of thumb for poster legibility is that when the poster is printed onto an 8.5 x 11 sheet, you should be able to read all the text.
Think of a poster like a visual representation of a research article. As such, posters typically include some of the following sections:
A second option for the final project is to create a digital data analysis dashboard with the flexdashboard package. This allows you to host your data analysis online and present it for public audiences.
Creating a flexdashboard requires using R Markdown to create a document. First, open RStudio and create a new R Markdown file. In the dialog box that opens, click “From Template.” If flexdashboard is installed, this will be one of the options. If it is not an option, go back and install the package before proceeding.
A new document will open with a basic layout for several charts in an untitled webpage. To turn the code into an html file, click “Knit” in RStudio and the compiled project will appear in the “Viewer” window. This can be expanded, hosted, or published directly from the viewer.
The basic template will look something like this.
Though not covered today, dashboards can also be made with Shiny in R.
The flexdashboard package follows much of the same syntax as R Markdown generally but has some unique features.
The layout of a dashboard can be thought of as a series of rows and columns. The layout is set by manually defining either the columns and rows with the following code. The width of each element can be defined or R will automatically divide the space based evenly based on the number of rows or columns.
Column {data-width=700}
-------------------------------------
Row
-------------------------------------Each column or row is populated by content blocks, dnoeted by a third
level heading ###. In context, this looks like this
Column {data-width=700}
-------------------------------------
### Block 1
### Block 2
Column {data-width=300}
-------------------------------------
### Block 1
Read more about layouts here.
R code is executed in chunks denoted by the following code, but uncommented.
# ```{r}
# ```General formatting is done in markdown. So the code…
- **bold**
- *italic*
> block quote…will produce the following:
block quote
The next section will walk you through how to create a sample dashboard using the Otter observation data from Stephens and Eckert (2019). Full citation: Tiffany Stephens and Ginny Eckert. 2019. Boat-based counts of sea otters at specific sites in Southeast Alaska. Knowledge Network for Biocomplexity. urn:uuid:b910f74b-171b-4d2b-b065-fb21823a8e84.
This dashboard uses example tables and figures from the R tutorials we have covered previously in this course.
Column {data-width=500}
-----------------------------------------------------------------------
### Chart A
Column {data-width=500}
-----------------------------------------------------------------------
### Chart B
### Chart C
### Chart D
Create your first chart in the left column of the flexdashboard. Remember to execute the R code with the right syntax.
otter <- read.csv("https://maddiebrown.github.io/ANTH630/data/sea_otter_counts_2017&2018_CLEANDATA.csv")notterpersite <- aggregate(n_otter ~ site_name, FUN = sum, data = otter)
notterpersite <- notterpersite[order(notterpersite$n_otter, decreasing = T), ]
top5 <- notterpersite[1:5, ]
top5$site_name <- reorder(top5$site_name, top5$n_otter)
p1 <- ggplot(top5, aes(x = site_name, y = n_otter)) + geom_col() + labs(x="Site name", y="Number of otters observed")
p1 + coord_flip()In addition to graphs, flexdashboards can also include text data, images, tables, and other types of content.
Text can be added to a flexdashboard simply by typing the text under the content block heading.
### About this dashboard
- This example dashboard uses otter observation data from Stephens and Eckert (2019).
- **Full citation:** Tiffany Stephens and Ginny Eckert. 2019. *Boat-based counts of sea otters at specific sites in Southeast Alaska*. Knowledge Network for Biocomplexity. [urn:uuid:b910f74b-171b-4d2b-b065-fb21823a8e84.](https://knb.ecoinformatics.org/view/urn%3Auuid%3Ab910f74b-171b-4d2b-b065-fb21823a8e84#urn%3Auuid%3A7eba259b-eeb5-4375-9596-e797bbb0b27d)Tables are created in flexdashboards using R Markdown syntax or packages designed to create tables.
With markdown, you define each cell of the table.
| | | |
|---|---|---|
| | | |
| | | |Each cell can be filled with content.
| test | A | B |
|------|---|---|
| 1 | 2 | 3 |
| 4 | 5 | 6 |For example, the code above generates the following table:
| test | A | B |
|---|---|---|
| 1 | 2 | 3 |
| 4 | 5 | 6 |
A helpful markdown table generator can be found here.
More information on tables in R Markdown can be found here.
kable() function from the knitr package,
replace Chart C with a table based on the following code: (note: you
may need to install and load the knitr package first)top_n(otter, 10, n_otter) %>% select(site_name, date_DDMMYY, year, n_otter) %>% arrange(desc(n_otter))
Uncomment the R code chunks to run this code.
### Top otter observations in a single day
#```{r}
#topotterdays <- top_n(otter, 10, n_otter) %>% select(site_name, date_DDMMYY, year, n_otter) %>% arrange(desc(n_otter))
#kable(topotterdays) # you can also directly input the code used to create topotterdays into the kable() function
#```
Value boxes are a great way to highlight important values or data. Value boxes can also be used as buttons linking to other webpages. The values of a value box can be static or update based on code running in the background. The color can be customized and many different types of icons can also be incorporated.
Depending on the version of R used, it seems like some of the modern versions of icons may not work seamlessley from certain sources. They could be added as html, but for the purposes of this tutorial, we will add icons directly into value boxes by name. Try searching for the older name for a particular icon if it doesn’t work on the first try. For ion icons, use this cheatsheet.
In the Chart D section, add a new value box for the total number of otters.
use the tally() function from tidyverse).Uncomment the r code lines.
### Chart D
#```{r}
totalotter<- otter %>% tally(n_otter) # 3188
valueBox(value=totalotter,"total otters observed",color="red", icon="ion-android-boat")
#```OR
### total otters observed
#```{r}
totalotter<- otter %>% tally(n_otter) # 3188
valueBox(3188, color="green", icon="fa-pencil")
#```Rather than stacking content blocks in columns, they can be included
as multiple pages accessed by tabs. A column can be turned into a tabset
by changing its attributes {} to
.{tabset}.
Column {.tabset}
-------------------------------------
### Chart 2
### Chart 3
Column {data-width=500 .tabset}
-----------------------------------------------------------------------At this point, your dashboard should look something like this
Flexdashboards can also be made with multiple pages. This is handy when the data story cannot be conveyed in a single page alone. The syntax for multiple pages is as follows:
Page 1
=====================================
Page 2
=====================================
Page 3
=====================================
R will use the page names as headings in the navigation bar. The navigation bar can also be customized by changing the order, spacing, icons, or text. Read more about this here
Uncomment the lines of R code to run the code below.
About
=====================================
### About this dashboard
- This example dashboard uses otter observation data from Stephens and Eckert (2019).
- **Full citation:** Tiffany Stephens and Ginny Eckert. 2019. *Boat-based counts of sea otters at specific sites in Southeast Alaska*. Knowledge Network for Biocomplexity. [urn:uuid:b910f74b-171b-4d2b-b065-fb21823a8e84.](https://knb.ecoinformatics.org/view/urn%3Auuid%3Ab910f74b-171b-4d2b-b065-fb21823a8e84#urn%3Auuid%3A7eba259b-eeb5-4375-9596-e797bbb0b27d)
Data analysis
=====================================
Column
---------------------------------------
### Top otter observation sites
#```{r}
notterpersite <- aggregate(n_otter ~ site_name, FUN = sum, data = otter)
notterpersite <- notterpersite[order(notterpersite$n_otter, decreasing = T), ]
top5 <- notterpersite[1:5, ]
top5$site_name <- reorder(top5$site_name, top5$n_otter)
p1 <- ggplot(top5, aes(x = site_name, y = n_otter)) + geom_col() + labs(x="Site name", y="Number of otters observed")
p1 + coord_flip()
#```
Column
---------------------------------------
### Top otter observations in a single day
#```{r}
#topotterdays <- top_n(otter, 10, n_otter) %>% select(site_name, date_DDMMYY, year, n_otter) %>% arrange(desc(n_otter))
kable(topotterdays)
#```
### Total otters observed
#```{r}
# total number of otters
#totalotter<- otter %>% tally(n_otter) # 3188
#valueBox(value=totalotter,"Total otters observed",color="red", icon="ion-android-boat")
#```Your result should look something like this.
Storyboards further extend the functionality of flexdashboards by adding an interactive menu slider for different subpages on a single site.
View an example storyboard with source code on the html widgets showcase.
A whole dashboard can be set as a storyboard in the document setup at the top of the script. In addition, a single page can be set as a storyboard in the same way we defined a column as a tabset.
The basic syntax of a storyboard page is to set each graphical
element with a third-level heading ###. Then R code defines
the large graphic of interest. Finally, each panel can be further
annotated with a descriptive sidebar following ***.
Page {.storyboard}
=====================================
### Panel 1
#```{r}
#```
***
Description
### Panel 2
#```{r}
#```
***
Description
library(leaflet)
leaflet(otter) %>%
addTiles() %>%
addCircleMarkers(lng=otter$longitude_E, lat=otter$latitude_N)Uncomment the R code below to run this.
Delete column headings.
Data analysis {.storyboard}
=====================================
### **Top otter observations:** Comparing the sites and dates with the highest number of otters observed.
#```{r}
notterpersite <- aggregate(n_otter ~ site_name, FUN = sum, data = otter)
notterpersite <- notterpersite[order(notterpersite$n_otter, decreasing = T), ]
top5 <- notterpersite[1:5, ]
top5$site_name <- reorder(top5$site_name, top5$n_otter)
p1 <- ggplot(top5, aes(x = site_name, y = n_otter)) + geom_col() + labs(x="Site name", y="Number of otters observed")
p1 + coord_flip()
#```
***
**Top otter observations in a single day**
#```{r}
topotterdays <- top_n(otter, 10, n_otter) %>% select(site_name, date_DDMMYY, year, n_otter) %>% arrange(desc(n_otter))
kable(topotterdays)
#```
### Total otters observed
#```{r}
# total number of otters
totalotter<- otter %>% tally(n_otter) # 3188
valueBox(value=totalotter,"Total otters observed",color="red", icon="ion-android-boat")
#```
***
Otter observations were made at a variety of sites over two years (2017 & 2018). The number to the left represents the total number of observations made.
### Map of otter sightings
#```{r}
library(leaflet)
leaflet(otter) %>%
addTiles() %>%
addCircleMarkers(lng=otter$longitude_E, lat=otter$latitude_N)
#```Dashboards can also be customized using a number of built-in themes. You can add some of these bootswatch themes in the heading section of your document. Not all of the themes are compatible with the version of bootstrap included in R Markdown. If you know html or css, there are ways to further customize the page. For now, we can use the built-in themes to change the style of the page.
Simply add a theme: argument to the yaml header at the
top of the script.
---
title: "Dashboard example"
output:
flexdashboard::flex_dashboard:
orientation: columns
vertical_layout: fill
theme: flatly
---
Your final dashboard should be customized and look something like this (though with a different theme and text).
We’ve only scratched the surface of all the different ways to customize and add content to flexdashboards. For further inspiration, check out the following pages.
Once you’ve created your dashboard you’ll want to share it with the world. Simply hitting knit creates a .html file in your working directory. This file can be emailed or otherwise shared as a standalone output. This file can also be hosted online with a platform of choice. This final section introduces two free platforms for quickly and easily hosting R Markdown data products.
RPubs is a great, quick way to host data analysis and dashboards. All you need is an account to get started and you can publish documents directly from RStudio. More information can be found on their Getting Started page. Note that all documents published to RPubs are public.
Github pages is another great option for hosting web pages and dashboards created with R. First you will need a github account and to enable github.io pages. Note that while you can create private repositories in github, the github.io pages are public. Additional information can be found in the links below.