Figure 1: Why R? 2018 conference banner used for social media promotion. The background displays banks of the Odra river in Wroclaw – the city of Poland where the conference was held.

Why R? 2018 conference

The primary purpose of the Why R? 2018 conference was to provide R programming language enthusiasts with an opportunity to meet and discuss experiences in R software development and analysis applications, for both academia and industry professionals. The event was held 2-5 August, 2018 in a city of Wroclaw, a strong academic and business center of Poland. The total of approximately 250 people from 6 countries attended the main conference event. Additionally, approximately 540 R users attended the pre-meetings in eleven cities across Europe (Figure 2).

Why R? 2018 conference is the continuation of the Why R?’s first edition that took place Sep 27-29, 2017 at the Warsaw University of Technology in Warsaw (Poland). Given the success of the first event, this year’s conference extended its program concept and scope; importantly, Why R? 2018 conference was held as international.

Figure 2: Locations and dates of the Why R? 2018 main conference event and 11 Why R?-branded pre-meetings.

Conference program

The format of the conference was aimed at exposing participants to recent developments in the R language, as well as a wide range of application examples. It consisted of workshops, invited talks, field-specific series of talks, lighting-talks, special interest groups, and a full-day programming hackathon.

The conference program had a strong focus on machine learning techniques and applications, with mlr (Bischl, M. Lang, L. Kotthoff, J. Schiffner, J. Richter, E. Studerus, G. Casalicchio, and Z. M. Jones 2016) R package – an interface to a large number of classification and regression methods – being emphasized in a number of presentations, as well as employed during workshops and the hackathon provided by the mlr team. The scope of conference program included statistical methodology, data visualization, R code performance, building products based on data analyses, and R’s role in academia / industry.

The event offered extensive networking opportunities. The cocktail party was held at the conference venue on the 2nd conference day. In addition, convenient location in the close proximity of the old town market square facilitated many informal gatherings that were happening each conference day.

Why R? Pre-meetings

The novel idea of pre-meetings has proved to be successful in popularizing Why R? conference in the international community of R users. Eleven pre-meetings took place in Czech Republic, Denmark, Germany, Poland, and Sweden in the run-up to the Why R? main event. The pre-meetings either constituted a part of another conference, one day-long workshop and discussion event, or a meeting of a local R user group.

As R provides a versatile framework for reproducible research in different scientific domains (Gentleman and D. Temple Lang 2007; Gandrud 2013; Leeper 2014; Liu and S. Pounds 2014; Rödiger, M. Burdukiewicz, K. A. Blagodatskikh, and P. Schierack 2015), we considered the Why R? pre-meetings as a great opportunity to convey and popularize R as an analytics tool in groups of professionals from different fields. The pre-meeting held at International Biotechnology Innovation Days (IBID), an open-access conference held 23-25 May, 2018 at the Brandenburg University of Technology Cottbus - Senftenberg (Senftenberg, Germany)¹ is an example where the R came in close contact with scientist from other domains. IBID brought together specialists and experts in the fields of bioanalytics, biomedical and translational research, autoimmune diagnostics, digitalization, and engineering; hence it posed an excellent platform to promote R and the Why R? 2018 conference.

Workshops

Why R? 2018 conference had a wide portfolio of workshops:

Maps in R by Piotr Sobczyk (OLX Group). Piotr showed how to create spatial data visualization efficiently in the R. He gave a plenty of tips to follow, pitfalls to avoid and a number of useful hacks. Starting from a basic plot function, he covered the usage of ggplot2 as well as R packages that use interactive javascript libraries to prepare data reports.
iDash - Make your R slides awesome with xaringan by Mikołaj Olszewski (iDash) and Mikołaj Bogucki (iDash). The workshop introduced the xaringan (Xie, C. T. Ekstrøm, D. Lang, G. Aden-Buie, O. P. B. C. in rmarkdown/templates/xaringan/resources/default.css), P. Schratz, and S. Lopp 2018) package – an alternative approach to preparing a slide deck. The xaringan package allows customizing each slide entirely and previewing slides dynamically in RStudio; moreover, the export of the slide deck (natively in HTML) to a pixel-perfect PDF is fairly easy. As xaringan also uses RMarkdown, it allows for reproducible results.
Jumping Rivers - Shiny Basics and Advanced Shiny by Roman Popat (Jumping Rivers). The instructor Roman Popat from Jumping Rivers conducted two workshops. In the first (Shiny Basics), he gave an introduction to creating interactive visualizations of data using Shiny. Here, participants learned how to use rmarkdown and htmlwidgets; input and output bindings to interact with R data structures; and input widgets and render functions to create complete page layouts using shiny and shiny dashboard. The advanced Shiny workshop explored how to add functionality to shiny apps using javascript packages and code. In particular, it was showed how one might deal with routines in a Shiny application that take a long time to run and how to provide a good experience for simultaneous users of an app. Finally, the instructor showed how to create a standalone web server API to the R code and how to integrate the use of it into a Shiny application using the plumber (Technology, LLC, J. Allen, F. van Dunné, S. Vandewoude, and S. Software (swagger-ui) 2018) package.
DALEX - Descriptive mAchine Learning EXplanations by Mateusz Staniak(Uniwersytet Wrocławski). THe workshop covered tools for exploration, validation, and explanation of complex machine learning models. The packages explored in this workshop include mlr (Bischl, M. Lang, L. Kotthoff, J. Schiffner, J. Richter, E. Studerus, G. Casalicchio, and Z. M. Jones 2016), DALEX (Biecek 2018), live (Staniak and P. Biecek 2018), FactorMerger (Sitko and P. Biecek 2017), archivist (Biecek and M. Kosinski 2017), pdp (Greenwell 2017) and ALEPlot (Apley 2018).
Constructing scales from survey questions by Tomasz Żółtak (Educational Research Institute in Warsaw, Poland). Tomasz showed how to create scales based on sets of categorical variables using Categorical Exploratory/Confirmatory Factor Analysis (CEFA / CCFA) and IRT models. He used models with bi-factor rotation to deal with different forms of asking questions and corrected for differences in a style of answering questions asked using a Likert scale. In addition, it was showed how to correct self-assessment knowledge/skill indicators using fake items.
From RS data to knowledge – Remote Sensing in R by Bartłomiej Kraszewski (Forest Research Institute, Poland). Remote sensing data from different sensors is a rich source of information for studying the natural environment, natural phenomena and monitoring some extreme phenomena, such as floods. Bartłomiej presented R language packages that can be used to work with remote sensing data. These included (a) for geographic information system analysis: rgdal (Bivand, T. Keitt, B. Rowlingson, E. Pebesma, M. Sumner, R. Hijmans, E. Rouault, F. Warmerdam, J. Ooms, and C. Rundel 2018), rgeos (Bivand, C. Rundel, E. Pebesma, R. Stuetz, K. O. Hufthammer, P. Giraudoux, M. Davis, and S. Santilli 2018) and sf (Pebesma, R. Bivand, E. Racine, M. Sumner, I. Cook, T. Keitt, R. Lovelace, H. Wickham, J. Ooms, and K. Müller 2018); (b) for raster data processing: raster (Hijmans, J. van Etten, J. Cheng, M. Mattiuzzi, M. Sumner, J. A. Greenberg, O. P. Lamigueiro, A. Bevan, E. B. Racine, A. Shortridge, and A. Ghosh 2017); (c) for Airborne LaserScanning data processing: the lidR (Roussel, D. A. R. the documentation), F. D. B. F. a. bugs improved catalog features), and A. S. M. I. lassnags) 2018) package.
Introduction to Deep Learning with Keras in R by Michał Maj (Appsilon Data Science). The workshop covered many important aspects of Deep Learning with the Keras in R, including sequential model building, performing data ingestion and using pre-trained models and performing fine-tuning. The keras (Allaire, F. Chollet, RStudio, Google, Y. Tang, D. Falbel, W. V. D. Bijl, and M. Studer 2018) R package was explored.

Invited talks

The invited talks topics included domain knowledge from statistics, computer science, natural sciences, and economics. The speakers list presents as follows:

Tomasz Niedzielski (University of Wroclaw): Forecasting streamflow using the HydroProg system developed in R,
Daria Szmurło (McKinsey & Company): The age of automation – What does it mean for data scientists?,
Agnieszka Suchwałko (Wroclaw University of Technology): Project evolution – from university to commerce,
Bernd Bischl (Ludwig-Maximilians-University of Munich): Machine learning in R,
Artur Suchwałko (QuantUp): A business view on predictive modeling: goals, assumptions, implementation,
Maciej Eder (Institute of Polish Language): New advances in text mining: exploring word embeddings,
Thomas Petzoldt (Dresden University of Technology): Simulation of dynamic models in R,
Leon Eyrich Jessen (Technical University of Denmark): Deep Learning with R using TensorFlow.

Special Interest Groups

Three Special Interest Groups were organized to facilitate topic-specific discussion between conference participants.

Diversity in Data Science, moderated by R-Ladies Warsaw, aimed to discuss boosting the diversity of R community and inspire members of affinity groups to pursue careers in data science.
The Career planning in data science, moderated by Artur Suchwałko (QuantUp) and Marcin Kosiński (Why R? Foundation), gave participants a chance to learn from experienced R enthusiasts about their career paths.
Teaching of data science, moderated by Leon Eyrich Jessen (Technical University of Denmark) and Stefan (Brandenburg Technical University Cottbus-Senftenberg), gathered data science experts from academia an industry to share their experiences and discuss challenges and solutions in teaching different concepts of data science.

Conference organizers

The quality of the scientific program of the conference was the achievement of Marcin Kosiński, Alicja Gosiewska, Aleksandra Grudziąż, Malte Grosser, Andrej-Nikolai Spiess, Przemysław Gagat, Joanna Szyda, Paweł Mackiewicz, Bartosz Sękiewicz, Przemysław Biecek, Piotr Sobczyk, Marta Karaś, Marcin Krzystanek, Marcin Łukaszewicz, Agnieszka Borsuk - De Moor, Jarosław Chilimoniuk, Michał Maj, and Michał Kurtys. The organization was in the hands of Michał Burdukiewicz (chair).

The organizers want to acknowledge R user groups from Berlin, Copenhagen, Cracow, Hamburg, Munich, Poznan, Prague, Stockholm, TriCity, Wroclaw, and Warsaw.

Acknowledgements

We would like to say thank you to all the sponsors, the University of Wrocław, Wrocław Center of Biotechnology Consortium, the local organizers of the pre-meetings, the mlr team, and student helpers.

Additional information

Why R? 2018 website http://whyr.pl/2018 Corporate sponsors: McKinsey & Company, Wrocław Center for Biotechnology, KRUK S.A., iDash s.c., R Consortium, WLOG Solutions, Jumping Rivers Ltd., RStudio, Inc., AnalyxGmbH, and Pearson IOKI.

Conference Report: Why R? 2018

Michał Burdukiewicz

Marta Karas

Leon Eyrich Jessen

Marcin Kosiński

Bernd Bischl

Stefan

2018-12-01