How to get into R

Alexander Matrunich
18 May 2017, R-Ladies Tbilisi

What are we going to do

  • Who am I?
  • Why R?
  • What is R?
    • Short R history
    • Basic elements of R installation
    • RStudio IDE
  • How to R?
  • Contacts

Alexander (Alex) Matrunich

Specialist degree in sociology at Pskov Volny Institute

Free open source enthusiast (GNU/Linux, LibreOffice, etc.)

Open data contributor (Wikipedia, OpenStreetMap)

Notable projects with R:

  • Opinion poll data processing and report generation
  • Exit-poll real-time data processing and report generation
  • R training for the UN FAO staff
  • World trade data analysis for the UN FAO

Why R?

R has overtaken SAS in job listings for "statistics"

Screenshoot by David Smith from Indeed Job Trends

R and Excel: difficulty vs complexity

Idea and plot by Gordon Shotwell

Average age in Swiss municipalities, 2015
Plot by Timo Grossenbacher. See notes by David Smith

Population structure of Japan

  • idbr package to get the US Census data
  • ggplot2 package to visualize
  • ggthemes package to apply The Economist style
  • animation package to create gif

See David Smith's post for details and link to source code

Where Europe lives, in 14 lines of R code

See details in David Smith's post

Where Europe lives, in 14 lines of R code

read_csv('GEOSTAT_grid_POP_1K_2011_V2_0_1.csv') %>%
  rbind(read_csv('JRC-GHSL_AIT-grid-POP_1K_2011.csv') %>%
          mutate(TOT_P_CON_DT='')) %>%
  mutate(lat = as.numeric(gsub('.*N([0-9]+)[EW].*', '\\1', GRD_ID))/100,
         lng = as.numeric(gsub('.*[EW]([0-9]+)', '\\1', GRD_ID)) *
           ifelse(gsub('.*([EW]).*', '\\1', GRD_ID) == 'W', -1, 1) / 100) %>%
  filter(lng > 25, lng < 60) %>%
  group_by(lat = round(lat, 1), lng = round(lng, 1)) %>%
  summarize(value = sum(TOT_P, na.rm = TRUE))  %>%
  ungroup() %>%
  complete(lat, lng) %>%
  ggplot(aes(lng, lat + 5*(value/max(value, na.rm = TRUE)))) +
  geom_line(size = 0.4, alpha = 0.8, color='#5A3E37', aes(group=lat), na.rm=TRUE) +
  ggthemes::theme_map() +
  coord_equal(0.9)

See details in David Smith's post

Twitter Faces

A collage of profile pictures of people who use the #rstats hashtag in their Twitter bio.

Created with R!

  1. Search users with the rtweet package.
  2. Download profile images with the httr package.
  3. Resize pictures and assemble the collage with magick package.

See details in the original post by Maëlle Salmon and a description in David Smith's post.

About ladies... R-ladies

Gender of developer See slides R Foundation Women in R Taskforce

What is R?

History of R

1976 - S language in Bell Labs by John Chambers

1993 - R - open source implementation of S by Ross Ihaka and Robert Gentleman at the University of Auckland

1997 - CRAN; part of the GNU Project

2000 - Version 1.0, suitable for production use

2004 - Version 2.0

2013 - Version 3.0. Values 231

2016 - R-Ladies Tbilisi founded

Base R Software

  1. Base R (written in C, Fortran and R)
  2. An R package:
    • Collection of R functions
    • Documentation (manuals and optionally vignettes)
    • Optionally data sets
    • Optionally inclusions of other programming languages (C++, Python, Java)
  3. CRAN - Comprehensive R Archive Network (10K packages January 2017)
  4. Every R user has:
    1. Base R installation
    2. System library of packages
    3. User's library of packages

Task views

plot of chunk unnamed-chunk-1

Integrated Development Environments for R

Free open source RStudio IDE:

  • Syntax highlighting, code completion, and auto indentation
  • Execute R code directly from the source editor
  • Quickly jump to function definitions
  • Easily manage multiple working directories using projects
  • Integrated R help and documentation
  • Interactive debugger to diagnose and fix errors quickly
  • Extensive package development tools

How to R?

How to start

  • Work in RStudio (if you are not an Emacs/Vim/… user)
  • Every job in separate directory (projects in RStudio)
  • Don't save and don't restore workspace image (settings in RStudio)
  • Follow code style
  • Restart R-session to avoid traces (Ctrl+Shift+F10 in RStudio)
  • Comment your code: why but not what

Errors

  • Read them!
  • Search in English
  • Use r, rstat, rlang, cran keywords
  • Short reporducible examples when asking for help

Storages of R-knowledge

https://insights.stackoverflow.com/trends?tags=r%2Cstatistics Stackoverflow Trends

  • Stackoverflow
  • Twitter: #rstat, Hadley Wickham, David Smith
  • Blogs: R-Bloggers

Contacts

Alexander (Alex) Matrunich

a@matrunich.com

Twitter: @matrunich