by Przemyslaw Biecek
The first meeting of R users in Poland took place in Wroclaw in 2008. It was a one-day conference with 27 participants and 6 talks.
Today, we have three large groups of R users in major Polish cities (according to meetup.com there is 640 users in SER – Warsaw, 235 in eRka – Cracow and 64 in PAZUR – Poznań). And there are also some data science groups that are full of R users anyway (450 users in Wroclaw and 115 in Łódź).
We have tried different forms of meetings. Some of them were working; some of them were not. Below I will summarize what we have learnt from them.
Today, the most common formula for our meetings is: two short talks (around 30 minutes) with 30 min break for pizza and networking. The main advantage of this formula is that it is very easy to organize. Just find a large room, order pizza or other snacks and find speakers. You can meet even without speakers, but with them the meeting is much nicer.
In Warsaw we have meetings every month. They usually start at 6pm. Currently, for meetings, we are using a class room at Warsaw Technical University. Previously we were meeting in pubs, coffee shops, other places, but finally it turns out that we need more space, good projector, good access to public transport, and we need place that rather quite silent (in coffee shop you sometimes compete with noise from coffee maker).
Among attenders there are students, some stuff (just few people) but majority of participants are graduates that are now working in data-oriented businesses. So 6pm is a good starting time since we can meet after work. We finish around 8pm – good time for after-party.
We also have organised hackatons (two in Warsaw, three in Cracow – watch the promo),
which were pretty cool, but such meetings require much more preparation and organization. Some of them resulted in interesting outcomes, like this diagram of flow between parties in Poland in previous cadence of Sejm. And there is fun in working with completely new people. After few hours it turns out that you can easily collaborate and you have learned a lot.
Talks are very diversified. Some of them are in Polish some in English, some are technical some are applied and some related to methodology. The 30 minutes per talk turns out to be enough to get people interested and ignite discussions.
Once we have tried a cinema–like meeting. Together (around 40 people) we were watching videos about deep networks, eating pizzas and sharing experience related to what we have heard. It wasn’t bad, but after all it is much better to have a live speaker.
After all I think that it is good to try different forms. It is interesting to see what is working and what is not.
The key ingredients are of course speakers. So I just would like to send thanks to our recent roster.
Karol Kubicki – Magic with pivotTable; Paweł Cejrowski – The package elastic; Zawistowski Paweł – Machine learning in production; Kornel Kiełczewski – From idea to model, What Could Possibly Go Wrong?; Markus Gesmann – Experience vs. Data; Emilia Kalarus – Non life insurance in R; Adam Wróbel – Stochastic mortality modelling; Piotr Migdal – Jupiter; Adolfo Alvarez – 5 lessons I have learned at Analyx; Wit Jakuczun – The protector – how to protect R code; Roger Bivand – Spatial data and R; Bartosz Meglicki – Machine learning and robotics; Wit Jakuczun – R in production; Paweł Chudzian – SVM in action; Piotr Dittwald – R and genomic rearrangements; Filip Grotkowski – R, robots and intelligent warehouses; Piotr Przybyła – Text mining of Sejm RP; Bogumił Kamiński – Integration of R and localsolver; Tomasz Żółtek – R and psychometrics; Paweł Teisseyre – Multilabel classification; Olga Mierzwa – price elasticity and the systemfit package; Teresa Ponikowska – R5 classes; Marek Gagolewski – the stringi package; Cezary Dendek – Shiny in buissness; Krzysztof Misztal – The package CEC (Cross Entropy Clustering); Zygmunt Zawadzki – R optimization; Piotr Ocalewicz – R and KNIME (video); Kamil Stanuch – Justin Bieber vs. The Beatles – text mining; Maciej Eder – The package Stylo (part 1); Ruprech von Waldenfels – Comparing the functional domain of linguistic elements across languages; Maciej Eder – The package Stylo (part 2); Jarosław Iwaniec – About hackathon. eRka and Onet.pl; Maciej Beręsewicz – RcppArmadillo; Alicja Szabelska – ampliQueso; Jakub Nowosad – machine learning with caret; Paweł Kleka – Item response theory in R, Mikołaj Bogucki and Krzysztof Jędrzejewski – Performance analysis of assessment content in an English language learning product; Łukasz Smaga – Functional data, Tomasz Górecki – Profiling in R; Other presentation from 13 PAZUR meetings
And we would like to thank our sponsors for they kind support (sometimes we get food and sometimes a good speaker): Revolution Analytics (now Microsoft) has supported us since 2009. Now, they are supporting all three polish R user groups!!!; datahero.tech; SmarterPoland.pl; WLOG Solutions;Codilime.com; Nationale-Nederlanden; Triple A – Risk Finance; INIME; Hub:raum; KoalaMetrics; Analyx.
The community is large and has big dreams. In 2014, we had three day Polish R conference called PAZUR. This year (October 12-14) we are going to organize European R users meeting [eRum] in Poznań. So be prepared to meet us at eRum!