An Introduction to R

• R is the new S
• Originally created to do stats and produce graphics
• It's a wholly extendable environment
• Growing quickly

• It's open source - try/use it for free
• Integrates with most data sources
• Can work inside other languages
• Customisable/extendable to your circumstances
• Really, really fast
• Some super snazzy functionality

• Written by stats bods so quite unique
• Stats emphasis can make it difficult to read documentation
• Myriad extensions can make it tough to know what to use
• It can be a bugger to google for at times

• Ad-hoc analysis
• Reproducible documents/reports
• Interactive reports
• Heavy duty number crunching

• Excel
  • PowerView etc
  • Reports for people to use
• Word
  • Reports on analysis and findings
• PowerPoint
  • Quick presentations
  • Presentations with data

• SAS / SSPS / SSAS
  • Data mining
  • Statistical analysis
• ETL
  • Connect to a lot of sources
  • In-memory and parallel processing for fast activities
• SSRS
  • Static reports
  • Exported/distributed reports

• Install R r-project.org
• Install Rstudio Rstudio.com <- Best IDE ever
• You're done

# Connection courtesy of Redgate bit.ly/1jrl7V6

# Activate the ODBC package for use
library("RODBC")

# Build a connection to the DB for reuse
azure <- odbcDriverConnect(
          "Driver={SQL Server Native Client 11.0};
          Server=mhknbn2kdz.database.windows.net;
          Database=AdventureWorks2012;
          Uid=sqlfamily;
          Pwd=sqlf@m1ly;")

# Store simple SELECT statement results
person <- sqlQuery(azure, 
         "SELECT TOP 10 * FROM [Person].[Person]")

# SQL statement used by some SSRS reports
sales <- sqlQuery(azure, "
SELECT 
soh.OrderDate,
t.Name as Territory,
cr.countryregioncode as Code,
cr.Name as Region,
sum(TotalDue) as Value,
count(*) as Volume
FROM [Sales].[SalesOrderHeader] soh
INNER JOIN [Sales].[SalesTerritory] t 
          on t.territoryid=soh.territoryid
INNER JOIN [Person].[CountryRegion] cr 
          on cr.countryregioncode=t.countryregioncode
WHERE  soh.OrderDate < '2008-07-01'
GROUP BY soh.OrderDate,
t.Name ,
cr.Name,
cr.countryregioncode")

Use data.table - feels like SQL and is REALLY fast

DT[ where, select | update, group by] [ having ] [ order by ] [ ]…[ ]

library("data.table")
library("zoo")
library("scales")

# summarise data
sales.sum <- data.table( sales )[, 
    list( Value = sum(Value), Volume = sum(Volume)), 
    by = list( YearQtr = as.yearqtr(OrderDate), 
        Region, Code)]

p<-ggplot( data = sales.sum, aes( x = YearQtr,  y = Value, group = Region, 
    colour = Region )) + geom_line() + 
    scale_x_yearqtr() 
p

p <- p + theme_minimal() + theme( legend.position = "bottom" ) + 
labs( x = "Order Quarter", y = "",   title = "Regional orders over time" ) + 
scale_y_continuous( labels = dollar ) + scale_x_yearqtr( format = "%YQ%q" ) 
p

library("ggthemes")
p<- p+facet_wrap( ~Region,  ncol = 3) + 
  theme_few()+
    theme( legend.position = "none",  axis.text.x = 
             element_text( angle = 45,   hjust = 1 )) 
p

p + aes(y=Volume) + scale_y_continuous() 

Insert or update

sqlSave(azure, sales.sum, salessumary)

Can do xlsx but I prefer keeping things in CSV

write.csv(sales, "sales.csv")

• This uses a package called shiny
• Allows local running of interactive reports
• Very easy to code (as it only uses R)
• Shiny Server available for hosting of reports
• Authentication systems available
• Examples

• This uses MikTeX (or LaTeX which is the platform independent name)
• Create an .Rnw file
• Include R code chunks
• Run knit2pdf command

knit2pdf(salesanalysis.Rnw)

• This uses markdown
• Create an .Rmd or .Rpres file
• Include R code chunks
• Run knit2html command
• Example = Supporting HTML file

knit2html(salesanalysis.Rmd)

• This uses markdown
• Create an .Rpres file
• Include R code chunks
• Use GUI to publish to Rpubs or save project to dropbox/github etc
• Example = this presentation

• R
• Rstudio
• Git (TFS integration)
• LaTeX
• Linux (I know!)
• Azure

• RODBC for a single layer of access
• data.table my go-to package for all data manipulation
  • data.table homepage
• zoo adds some excellent date/time functionality
• ggplot2 my go-to charting package
  • ggplot2 homepage

• xtable tables in documentation
• knitr documentation
• shiny interactive reports

• testthat unit testing
• devtools general package dev

• tryR - online beginners course
• Advanced R programming
• rseek.org - search engine
• stackoverflow
• r-bloggers.com - consolidated blogs
• ggplot2 book
• Web application development with R and Shiny
• writelatex.com - online document dev
• LaTeX wiki