class: center, middle, inverse, title-slide # useR!2017 ## Warwick R User Group ### Ella Kaye & David Selby ### 20^th July 2017 --- background-image: url(images/user-sign.jpg) background-size: cover ??? ## Contents 1. useR!2017 conference - event itself (DS) - inclusivity of conference (EK) - community (EK) 2. Packages - 20 years of CRAN (DS) - package highlights (DS/DV) - package highlights (EK) 3. future plans - within R / R community (DS/EK) - for WRUG (DS) NB: more photos on the conference web site. --- background-image: url(https://www.user2017.brussels/uploads/_full/deel-2-web-13.jpg) background-size: cover ??? Breakout room --- background-image: url(https://www.user2017.brussels/uploads/_full/sfeer-dag-3-87.jpg) background-size: cover ??? Breakout room full --- background-image: url(https://www.user2017.brussels/uploads/_full/deel-2-web-78.jpg) background-size: cover ??? Plenary room --- background-image: url(https://www.user2017.brussels/uploads/_full/dag-2-sfeer-37.jpg) background-size: cover ??? Plenary room full --- background-image: url(images/atomium.jpg) background-size: cover --- background-image: url(images/dinner.jpg) background-size: cover --- background-image: url(images/tea.png) background-size: contain background-position: 50% 15% --- class: fullpage .pull-left[ ### Welcoming and inclusive - newbies session - 25 diversity scholarships - transport for people with impairments - breast-feeding area - quiet zone - reasonable fees for students - childcare - Heather Turner / Forwards / R Foundation ] ??? EK: Welcoming and inclusive / People come -- .pull-right[ ### People come &nbsp; - 1161 attendees - 54 countries - 941 litres of beer ] --- background-image: url(https://image.slidesharecdn.com/rthenandnow-microsoftsponsortalk-170705092253/95/r-then-and-now-9-638.jpg?cb=1499246642) background-size: contain ??? EK: R ecosystem --- class: center, bottom, section background-image: url(https://pbs.twimg.com/media/DD98lZPW0AEaXP0.jpg:large) background-size: contain background-position: 50% 0% [RLadies.org](https://rladies.org) .secondary[|] [@RLadiesGlobal](https://twitter.com/RLadiesGlobal) .secondary[|] [@RLadiesLondon](https://twitter.com/RLadiesLondon) ??? EK: R-Ladies --- class: middle # .primary[R-Ladies .secondary[`<-`]] Worldwide organization that promotes gender diversity in the R community via meetups and mentorship in a friendly and safe environment 20+ countries .secondary[|] 45+ cities .secondary[|] 6000+ members ??? EK: R-Ladies --- # Collaborative coding.secondary[\*] - Step-by-step on contributing to package development, especially on GitHub - If you're here, you're ready to contribute to R!  .footnote[.secondary[\*] _Charlotte Wickham, Oregon State University & DataCamp_ ([slides][CWslides])] [CWslides]: http://cwick.co.nz/talks/collab-code-user17/ ??? EK: collaborative coding --- layout: true # 20 years of CRAN.secondary[\*] --- <dl> <dt>1976</dt> <dd>“At the beginning was S”</dd> <dt>1992</dt> <dd>R development begins</dd> <dt>1997</dt> <dd>R Core team and CRAN founded</dd> <dt>2000 (February 29)</dt> <dd>R-1.0.0 released (compatible with S3)</dd> <dt>2013</dt> <dd>R-3.0.0 released</dd> </dl> .footnote[.secondary[\*] _Uwe Ligges, R Core_ ([video][cran])] [cran]: https://channel9.msdn.com/Events/useR-international-R-User-conferences/useR-International-R-User-2017-Conference/KEYNOTE-20-years-of-CRAN ??? DAS: 20 years of CRAN --- **11,100** packages depend on other packages that depend on other packages that depend on other packages... Packages with >1000 reverse dependencies: - knitr - testthat - Rcpp - MASS .comment[\>75% of all packages] - Matrix .comment[\>75% of all packages] - survival .comment[\>75% of all packages] .footnote[.secondary[\*] _Uwe Ligges, R Core_ ([video][cran])] --- layout: false class: fullpage, middle  ??? DAS: Tweet about Sparc (Solaris) hardware --- layout: true # New R documentation.secondary[\*] ??? DAS - R Documentation Task Force - dynamic documentation - **parsetools** and **documentation** packages --- Documentation stored/manipulated as S4 objects  .footnote[.secondary[\*] _Andrew Redd, University of Utah_ ([video][Rdocs])] [Rdocs]: https://channel9.msdn.com/Events/useR-international-R-User-conferences/useR-International-R-User-2017-Conference/Updates-to-the-Documentation-System-for-R --- ```r hw <- function( greet = "Hello" #< A greeting , who = "World" #< who to greet. ){ #! A more complicated hello world print(paste(greet, who)) #< Called for the side effect of printing, #^ but returns the pasted arguments invisibly. } ``` Implicit unit testing ```r if (FALSE) {#@testing ... } ``` .footnote[.secondary[\*] _Andrew Redd, University of Utah_ ([video][Rdocs])] --- layout: false # Teaching data science to new useRs.secondary[\*] Teaching data science to first-year undergraduates at Duke - “**tidyverse** first” - RStudio instances - Reproducibility with R Markdown - Version control with GitHub - Supplement with DataCamp and [**learnr**] package &nbsp; .footnote[.secondary[\*] _Mine Çetinkaya-Rundel, Duke University & RStudio_ ([video][mcr link])] [**learnr**]: https://rstudio.github.io/learnr/ [mcr link]: https://channel9.msdn.com/Events/useR-international-R-User-conferences/useR-International-R-User-2017-Conference/KEYNOTE-Teaching-data-science-to-new-useRs ??? DAS: teaching new useRs --- class: center, middle, inverse, section # Packages --- class: fullpage ## [**poissontris**]: Shiny game  [**poissontris**]: https://github.com/openanalytics/poissontris ??? DAS Maybe do a live demo at the start or end. --- layout: true class: fullpage ## [**miner**](https://ropenscilabs.github.io/miner_book/): R interface for Minecraft ---  --- count: false  --- count: false  --- layout: false class: fullpage ## [**papr**]: Tinder for pre-prints - By Jeff Leek ([simplystatistics.org](https://simplystatistics.org)) - Supports [bioRxiv](http://www.biorxiv.org/) (maybe arXiv in the future)  [**papr**]: https://jhubiostatistics.shinyapps.io/papr/ ??? DAS This is not meant to be a serious project; rather it demonstrates Johns Hopkins Data Science Lab's ability to prototype data apps. Blog post: https://simplystatistics.org/2016/10/03/papr/ --- background-image: url(http://tidytextmining.com/images/cover.png) background-size: 25% background-position: 85% 80% ## Text mining, the tidy way.secondary[\*] 1. Each variable is a column 2. Each observation is a row 3. Each type of observation is a table .secondary[`->>`] one *token* per row Tidy text packages: - **tidytext** - **manifestoR** - **cleanNLP** - **quanteda** .footnote[.secondary[*] _Julia Silge, Stack Overflow_ ([video][tidytext video])] [tidytext video]: https://channel9.msdn.com/Events/useR-international-R-User-conferences/useR-International-R-User-2017-Conference/Text-mining-the-tidy-way ??? DAS e.g. sentiment analysis of Jane Austen novels --- class: fullpage ## Programming with the tidyverse.secondary[\*] .pull-left[ ```r band_members ``` ``` ## # A tibble: 3 x 2 ## name band ## <chr> <chr> ## 1 Mick Stones ## 2 John Beatles ## 3 Paul Beatles ``` ```r x <- 'John' filter(band_members, name == x) ``` ``` ## # A tibble: 1 x 2 ## name band ## <chr> <chr> ## 1 John Beatles ``` ] -- .pull-right[ ```r y <- 'name' filter(band_members, y == x) # Nothing found! ``` ```r z <- rlang::sym('name') filter(band_members, (!! z) == x) ``` ``` ## # A tibble: 1 x 2 ## name band ## <chr> <chr> ## 1 John Beatles ``` &nbsp; .footnote[.secondary[*] _Lionel Henry, RStudio_ ([slides][LHslides])] ] [LHslides]: https://schd.ws/hosted_files/user2017/43/tidyeval-user.pdf --- class: fullpage ## Using RStan in packages.secondary[\*] - **rstantools** - provides `rstan_package_skeleton` - **rstanarm**: RStan for **a**pplied **r**egression **m**odels - `stan_lm`, `stan_aov` for Anova models - `stan_glm` for generalised linear models - `stan_lmer`, `stan_glmer` for mixed-effects models .footnote[.secondary[\*] _Ben Goodrich, Columbia University_ ([video][BG talk])] -- ```r glm(y ~ roach1 + treatment + senior, offset = log(exposure2), data = roaches, family = poisson) stan_glm(y ~ roach1 + treatment + senior, offset = log(exposure2), data = roaches, family = poisson, prior = normal(0,2.5), prior_intercept = normal(0,5), chains = CHAINS, cores = CORES, seed = SEED) ``` [BG talk]: https://channel9.msdn.com/Events/useR-international-R-User-conferences/useR-International-R-User-2017-Conference/How-to-Use-RStan-to-Estimate-Models-in-External-R-Packages ??? DAS R package **rstantools** provides utilities to make it easier to make your own packages that use Stan. R package **rstanarm** is an example of a package that uses these tools. It supplies Bayesian versions of common regression models, with a familiar interface similar to the **stats** package. --- class: section background-image: url(http://wallpapercave.com/wp/9dsOTpr.png) background-size: cover ??? EK --- ## [**narnia**]: consistent exploration of missing data.secondary[\*]  .footnote[.secondary[\*] _Nicholas Tierney, Monash University_ ([web site][NT])] [NT]: http://narnia.njtierney.com/ [**narnia**]: https://github.com/njtierney/narnia ??? EK --- ## Other cool packages/functions for missing data &nbsp; - [visdat](https://github.com/njtierney/visdat) .comment[visualise missing data] - [simputation](https://github.com/markvanderloo/simputation) .comment[make imputation simple] - [`mice::ampute`](https://cran.r-project.org/web/packages/mice/vignettes/ampute.html) .comment[simulate missing values] ??? EK --- ## Top tips ### Speed up your code Add `ByteCompile: yes` to your package `DESCRIPTION` ### Interactive comments ```r ggplot(iris, aes(Petal.Width, Petal.Length)) + geom_point() + labs( x = "Petal width" , y = "Petal length" # , title = "Fisher's iris data set" ) ``` --- # Resources - These slides [warwick.ac.uk/wrug](http://warwick.ac.uk/wrug) - Videos of _useR!2017_ talks.secondary[*] [bit.ly/2sUxtPJ](http://bit.ly/2sUxtPJ) - [R-Consortium.org](https://www.r-consortium.org/) - [RLadies.org](https://rladies.org/) - [satRdays.org](http://satrdays.org/) &nbsp; .footnote[.secondary[*] Dootika's is up! Ella's talk isn't there (yet)] --- class: center, middle, inverse, section # FutuRe --- background-image: url(images/brussels2.jpg) background-size: cover --- ## R User Groups - Bay Area User Group - format: mix of 3—4 talks of different lengths - 12-min lightning talks - 20/40-min presentations - no permanent venue - satRdays - R Consortium - Meetup.com - eRum / useR!2018 - Online / recorded meetings --- # Warwick R User Group - Now on [Meetup.com/Warwick-useRs](https://www.meetup.com/Warwick-useRs/) - R Consortium funding - Archived material at [warwick.ac.uk/wrug](http://warwick.ac.uk/wrug) - Evening/weekend events - Meeting venues - Online / recorded meetings --- ## Upcoming R events <dl> <dt>August</dt> <dd>Next WRUG meeting?</dd> <dt>11–15 September</dt> <dd> <a href="http://warwick.ac.uk/wdsi/vacationschool2017"> WDSI/IAS Vacation School </a> <br/> <em>Principles and Practice of Data Analysis for Reproducible Research in R</em></dd> <dt>12–14 September</dt> <dd><a href="https://earlconf.com">EARL conference</a> (London)<br/> <em>Enterprise Applications of the R Language</em></dd> </dl> --- class: section, inverse, middle, center # Thank you!