List of videos

"Building a full-stack web application with Python, NPM, Webpack and React [EuroPython 2017 - Talk - 2017-07-10 - Anfiteatro 1] [Rimini, Italy] Creating full-stack applications with Python, NodeJS and React can seem daunting at first. Having made many variations of these, I will show you the ropes, so you too can discover that it is in fact easy to get going. In this talk you will learn to create a full-stack web application in Python, with a Nodejs and React front-end. I will provide you with an easy-to- follow walkthrough of the process, and you’ll exit this talk feeling confident that you can now create your own full-stack web application. License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2017.europython.eu/en/speaker-release-agreement/

Opening session [EuroPython 2018 - - 2018-07-25 - Smarkets] [Edinburgh, UK] License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/

Die Threads [EuroPython 2018 - Keynote - 2018-07-25 - Smarkets] [Edinburgh, UK] By David Beazley In the brave new world of async, threads are now a thing of the past. Or are they not? License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/

ETL pipeline to achieve reliability at scale [EuroPython 2018 - Talk - 2018-07-25 - Fintry [PyData]] [Edinburgh, UK] By Isabel Lopez In an online betting exchange, thousands of money related transactions are generated per minute. This data flow transforms a common and, in general, tedious task such as accounting into an interesting big data engineering problem. At Smarkets, accounting reports serve two main purposes: housekeeping of our financial operations and documentation for the relevant regulation authorities. In both cases, reliability and accuracy are crucial in the final result. The fact that these reports are generated daily, the need to cope with failure when retrieving data from previous days, and the fast growing transaction volume obsoleted the original accounting system and required a new pipeline that could scale. This talk presents the ETL pipeline designed to meet the constraints highlighted above, and explains the motivations behind the tech stack chosen for the job, which includes Python3, Luigi and Spark among others. These topics will be covered by describing the main technical problems solved with our design: - Fault tolerance and reliability, i.e ability to identify faulty steps and only rerun those instead of the whole pipeline. - Fast input/output. - Fast computations. License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/

Hacking Reinforcement Learning [EuroPython 2018 - Talk - 2018-07-25 - Fintry [PyData]] [Edinburgh, UK] By Guillem Duran Creating huge datasets of top performing examples for Reinforcement Learning (RL) has always been tricky, but if we allow ourselves to cheat a bit it can be done very easily. During this talk, I will present a new family of algorithms that allow to efficiently generate very high quality samples for any known RL environment. This new generation of planning algorithms achieves a performance which is several orders of magnitude higher than any other existing alternative, while offering linear time complexity and good scalability. This talk will be a practical example of how we can use new tools for hacking any reinforcement learning environment, and make it generate superhuman level games. Hacking RL, as any other hacking process will be divided in four phases: During information gathering, I will briefly explain what are the main ideas behind Reinforcement Learning. I will also talk about how our theory (FractalAI) came to be, and what are the fundamental concepts behind it. We will find an attack vector against the environment API, and explain how it can be exploited. I will explain the fundamental concepts needed to build a new generation of exploits, that will allow us to have complete control over the data the environment produces. This is the time to test the new exploits and to show a proof of concept. We will exploit the attack vector to gain access to the environment. Using only a laptop I will show how it is possible to sample data which surpasses human performance way faster than real time. Once we have gained control of the environment, we will measure how well the exploits work, and how well the techniques presented can generalize to other types of environments. I want the talk to be as simple and fast as possible, with a lot of graphical examples, videos, and a Jupyter notebook. The Q&A session is the time to apply some social engineering to get me to talk about the details that you find more interesting. I have prepared additional material covering the most common questions and concerns, but feel free to ask whatever you want, I love challenging questions ;) License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/

All You Need is Pandas: Unexpected Success Stories [EuroPython 2018 - Talk - 2018-07-25 - Fintry [PyData]] [Edinburgh, UK] By Dimiter Naydenov Learning to use the awesome Pandas toolkit helped me immensely in lots of ways. Finding novel, efficient solutions to complex day-to-day problems with Pandas not only saves time, but can be fun and rewarding experience. In this talk I'll present use cases I had to solve, but the "traditional" approach proved tough and/or otherwise frustrating implement nicely. Since I was just starting to learn Pandas, decided to try an alternative solution with it. What I learned changed the way I think about data processing with Python, and it only got better since! The use cases deals with extracting pen strokes from handwritten SVG samples, and recomposing them into reusable letters and numbers. When you need to compare each stroke to all others, often more than once, resulted in inefficient, slow, and hard to maintain code. Even a naive Pandas approach with loops helped to reduce both the memory footprint, and improve the performance considerably! Improving the implementation further, vectorizing inner loops, and taking advantage of multi-index operations, I managed to get the same results, using less memory and a lot faster (by orders of magnitude). License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/

How is python used in biomolecular sciences? [EuroPython 2018 - Talk - 2018-07-25 - Fintry [PyData]] [Edinburgh, UK] By Antonia Mey In the last ten years scientists working on computational problems involving proteins and other small molecules have largely moved to using python when developing simulation and data analysis tools allowing for a fast prototyping and development of new ideas. One big challenge is dealing with the compatibility of different tools and using these to create very complex adaptive, yet robust workflows in order to be able to guide cutting edge experiments e.g. predicting how well a small drug like molecule can bind to a protein that could serve as a target for a new drug. The talk will give a gentle introduction to what kind of python related tools are available in the field of computational molecular biology, how they are used, and what kind of complex workflows scientist have to solve. I will then introduce BioSimSpace an open source python library and flagship project of the CCPBioSim consortium in the UK, which provides a common API to avoid having to learn many individual tools facing compatibility and dependency challenges allowing scientists to focus on the scientific question at hand and not solving programming challenges. BioSimSpace allows fast and interoperable building of workflow components (nodes) for bimolecular problems, which can easily be used on a variety of different computational resources. In particular I will introduce the cloud facilities available for fast prototyping using a Jupyter notebook interface. License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/

What's new in Python 3.7 [EuroPython 2018 - Talk - 2018-07-25 - Smarkets] [Edinburgh, UK] By Stephane Wirtel Scheduled for release in mid-June before the conference, Python 3.7 is shaping up to be a feature-packed release! This talk will cover all the new features of Python 3.7, including the Data Classes and the Context Variables for the asynchronous programming with asyncio. License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/

Adventures in compatibility: emulating CPython's C API in PyPy [EuroPython 2018 - Talk - 2018-07-25 - Smarkets] [Edinburgh, UK] By Ronan Lamy PyPy is a fast and compliant implementation of Python. In other words, it's an interpreter for the Python language that can act as a full replacement for the reference interpreter, CPython. It's optimised to enable efficient just-in-time compilation of Python code to machine code, and has releases matching versions 2.7, 3.5 and soon(ish) 3.6. The PyPy project also developed cffi as a clean and efficient way of interfacing with C code. However, many libraries in the Python ecosystem are implemented as C extensions, which target CPython's C API. Many others use Cython, which builds C extensions under the hood. Therefore, PyPy needs an emulation layer for the C API. This emulation needs to bridge the differences between the implementation languages and the object models of CPython and PyPy. The solution is called cpyext. It's implemented in a mixture of RPython and C, with most of the API functions and macros implemented in RPython. cpyext exposes PyObjects to the extension code that appear similar to CPython objects (as long as extension writers stay within the fuzzily defined boundaries of the public API) but are merely 'shadows' of the real PyPy objects. After a brief presentation of PyPy, its goals, and its current statuts and roadmap, this talk will dive into the vexed topic of its handling of C extensions. By the end of it, the audience should understand the operating principles of cpyext and have a clearer understanding of what happens when you install and run numpy, for instance, on top of PyPy. Some basic familiarity with CPython internals and how C extensions are made will be assumed. License: This video is licensed under the CC BY-NC-SA 3.0 license: https://creativecommons.org/licenses/by-nc-sa/3.0/ Please see our speaker release agreement for details: https://ep2018.europython.eu/en/speaker-release-agreement/