YAML

The YAML is the first lines of code telling how your document will be rendered. It lies in the top of your document between 3 dashed lines.

---
title: "R nanocourse 4: Dynamic Documents"
author: "Thierry Warin"
date: "12/02/2020"
output: 
  html_document:
    toc: yes
    toc_depth:3
  pdf_document:
    toc: yes
    toc_depth:3
---

In the output section, you have several options:

• toc: yes/no, which enable/disable the table of content
• toc_depth: number, which set the depth of the table of content

You can set a specific date to your document, but also change it so that it will render the actual date of rendering. For example, you are writing a document that will be compiled in three months, the date that will be shown will be the actual date. To do so, you can enter a command in R code that will seek for the actual date of your console, using the format() function.

date: `r format(Sys.time(), '%d %B, %Y')`

Task:

• Create a new document (.Rmd): File > New File > R Markdown...
  
• Set your YAML settings according to the date of knitting

R chunk settings

Every line of R code has to be confined between dashed lines for the RStudio console to interpret them.

However, it is possible to set different parameters in order to render different outputs for each R chunk. These settings have to specified in the first line of the R chunk, such as:

• echo = FALSE/TRUE: if FALSE, the code will not been shown
• warning = FALSE/TRUE: if FALSE, warnings will not been shown
• message = FALSE/TRUE: if FALSE, messages generated by the code will not been shown
• fig.align = 'center'/'left'/'right': will align the figure generated depending on the setting

# Loading packages
library(gsheet)
library(dplyr)

# URL of the UNIDO dataset
gs15x <- "https://docs.google.com/spreadsheets/d/1aTJFKmkH2oxYcg0aiWeMAttGWKdM1u2KS5OyIlUkI6Q/edit?usp=sharing"

# Using the gsheet2tbl function to import the UNIDO dataset into the RStudio console
dataUnido <- gsheet2tbl(gs15x)

# Transform variables into numeric values
dataUnido$Value <- as.numeric(dataUnido$Value)
dataUnido$Tablecode <- as.numeric(dataUnido$Tablecode)
dataUnido$CountryCode <- as.numeric(dataUnido$CountryCode)
dataUnido$Year <- as.numeric(dataUnido$Year)
dataUnido$IsicCode <- as.numeric(dataUnido$IsicCode)
dataUnido$Unit <- NULL

# Subset concerning only data for the IsicCode = 1542
dataUnidoSubset <- filter(dataUnido, IsicCode == 1542)

Number of employees

Task:

• Select a subset of the dataset regarding only the number of employees (Tablecode == 04)
  
• Provide for 2010 a ranking of the country with the most important number of employees
  
• Visualize and compare the top 7 countries in terms of employees

# Data regarding the number of employees
dataEmployees <- filter(dataUnidoSubset, Tablecode == 4)

# Data regarding 2010
dataEmployees2010 <- filter(dataEmployees, Year == 2010)

# List the 10 most important countries in terms of employees in 2010
ranking <- arrange(dataEmployees2010, desc(Value))

The list of the top 7 countries in terms of employees in the Sugar industry in 2010 are :

head(ranking, n=7)

Now, let's visualize these data in a bar chart.

library(ggplot2)
library(ggthemes)
library(reshape2)

# Transform the column 'CountryCode' in a factor type
ranking$CountryCode <- as.factor(ranking$CountryCode)

# Produce a bar chart
ggplot(data = ranking[1:7,], aes(x = CountryCode, y = Value, fill = CountryCode)) + 
  geom_bar(stat = "identity", width = 0.5, position = "dodge")  +  
  ylab("Number of employees")  +
  xlab("") +
  guides(col = guide_legend(row = 1)) +
  theme_hc() +
  scale_fill_brewer(direction = -1)

So the most important countries in terms of employees in the sugar industry in 2010 are:

• China
• Russia
• Mexico
• Iran
• Vietnam
• Ukraine
• Colombia

Number of establishments

Based on previous results, we would like to observe the evolution of the number of establishments in the top 3 countries as of 2010 in terms of employees (i.e. China, Russia, Mexico).

Task:

• From the sugar industry dataset, select data corresponding to these 3 countries for all available years
  
• Visualize the evolution of the number of establishments through time

# Subset of the dataEmployees dataframe concerning only the three selected countries
dataEmployeesCountries <- filter(dataEmployees, CountryCode == 156 | CountryCode == 643 | CountryCode == 484)

# Transform the column 'CountryCode' in a factor type
dataEmployeesCountries$CountryCode <- as.factor(dataEmployeesCountries$CountryCode)

# Produce a line chart
ggplot(data = dataEmployeesCountries, aes(x = Year, y = Value, color = CountryCode)) +
  geom_line()  + 
  ylab("")  +
  xlab("") +
  geom_smooth(span = 0.8) +
  ggtitle("") +
  theme_hc() +
  scale_color_brewer(direction = -1) +
  guides(fill=FALSE) +
  geom_point(colour = "blue", size = 2,shape = 22)

PDF / HTML / doc

Now that your analysis has been completed, you can export your document in different format from the same RMarkdown file. To do so, click on the arrow on the right of the "Knit HTML" buttom and select the appropriate format (HTML, PDF, doc).

Task:

• Render your RMarkdown document into a PDF file
  
• Render your RMarkdown document into a HTML file

Ioslide / Beamer presentation

From the same RMarkdown file, it is possible to generate a "powerpoint" presentation. To do so, please consider the following instructions. First, you have to change the YAML options: in the output field, insert:

• output: beamer_presentation (powerpoint)
  
• output: ioslides_presentation (interactive presentation)

Secondly, a proper typology has to be adopted:

• For a first-level slide, insert "#" before the title
• For a second-level slide, insert "##" before the title

Task:

• Open a new RMarkdown document
  
• Select the code corresponding to the industrial analysis
  
• Create an ioslide/beamer document
  
• Showcase your results in the two different formats

Resources

For more on the RMarkdown syntax, please refer to:

• https://www.rstudio.com/wp-content/uploads/2015/03/rmarkdown-reference.pdf

Packages

• ggplot2: H. Wickham & W. Chang
• ggthemes: J. Arnold et al.
• reshape2: H. Wickham
• gsheet: M. Conway
• dplyr: H. Wickham & R. Francois
• tidyr: H. Wickham