Julia prog language

Julia Users Group - Berlin


Julia Users Group - Berlin had its first get together in 2015. The discussion organising the first meetup can be found on the Julia Users Google group forum.

Following the 2016 PyData conference in Berlin a lot of interest was expressed in meeting more regularly to discuss Julia-related topics. Since then we've been trying to co-ordinate a bi-monthly meet-up of Julia users in Berlin.

Previous meetings


Location: tuesday coworking (Feurigstr. 51)
Topic: Lorenzo talked about nonsmooth optimization using proximal algorithms in Julia, in particular his work on the packages ProximalOperators.jl and ProximalAlgorithms.jl. The first one implements first order primitives for a variety of functions that can be used to express problems, the second one relies on the first and contains generic implementations of various well-known iterative methods in convex optimization. He also told us of his plans to develop a generic mathematical modelling language for numerical optimisation tasks.


Location: The Factory Berlin (Rheinsberger Str. 76/77)
Topic: David gave an overall introduction to the JuliaGPU project, with a focus on OpenCL. Simon followed this with an introduction to the GPUArrays interface and CLArrays, which he has worked on, as simple ways to offload your matrix arithmetic to the GPU.


Location: The Factory Berlin (Rheinsberger Str. 76/77)
Topic: Dirk presented his work on TransitmapSolver.jl, which is used to draw transit maps automatically. Simon gave us an update on GLVisualize.


Location: PyData Berlin, 2017
Topic: David and Robert gave a workshop on Julia for Data Science. Hector gave a talk on his Shearlets toolbox.


Location: The Factory Berlin (Rheinsberger Str. 76/77)
Topic: Robert presented an overview of optimisation tools which fall under the JuliaOpt umbrella. Following this he went into detail on JuMP, a package to which he is a contributor.


Location: The Factory Berlin (Rheinsberger Str. 76/77)
Topic: Hector spoke about using Julia for mathematical applications, using his Julia packages for Wavelets and Shearlets, in a talk entitled Filter-based implementation of Fast Wavelet and Shearlet transform in Julia. We also briefly discussed our plans for hosting an Introduction to Julia workshop aimed at numerically literate users. And we had an informal overview discussion of parallelisation using Julia.


Location: The Factory Berlin (Rheinsberger Str. 76/77)
Topic: Simon introduced GLVisualize meanwhile Cristina demonstrated Molcajete.jl a program which she wrote for accessing the Meet-up API in order to schedule non-clashing meet-ups. We also had our first disucssion of plans for hosting an introduction to Julia workshop.


Location: The Factory Berlin (Rheinsberger Str. 76/77)
Topic: Robert presented the Plots package and Vinny introduced us to some machine learning applications and embedding in a jupyter dashboard.


Location: The Factory Berlin (Rheinsberger Str. 76/77)
Topic: David presented a neuronal network simulator, written entirely in Julia. The focus was on what design choices were enforced/enabled by the use of Julia.


Location: St. Oberholz (Rosenthaler Platz)
Topic: Our kick-off meeting, following PyData 2016. Miles Lubin, co-author of JuMP library and a founder of JuliaOpt, was visiting and attended. We discussed a number of topics about the state of Julia, the development behind the scenes, of course some numerical optimisation topics, and the organisation of the meet-up.


Location: St. Oberholz (Rosenthaler Platz)
Topic: Our very first meet-up in Berlin. Viral Shah was in Germany on business and attended. Simon Danisch also made the trip to Berlin, and has subsequently moved here to join us more permanently. We had a great night, but the reality of organising a regular meet-up with so few members meant that it took another 15 months before the next meeting occured.