PyCon.DE 2017
as the organisers put it
PyCon.DE 2017 Karlsruhe - is the traditional meeting point for people who work and play with Python. Hosted in Germany and held in the English language, the conference welcomes Python professionals and enthusiasts from Central Europe and around the world.
| title | author | length |
|---|---|---|
| Neutrinos: who are you and if yes how many? | Susanne Mertens | 57m |
Direct measurement of the neutrino mass via a 70 m tritium spectrometer claims sensitivity beyond prior limits and an additional search for a sterile neutrino… | ||
| Dask: Next Steps in Parallel Python | Matthew Rocklin | 01h 03m |
Dask demonstrates running familiar pandas code on a cluster by breaking a 20 gigabytes on disk dataset into ~300 blocks, executing on a 64 cores cluster with a… | ||
| AI: Differentiating Hype and Real Value | Michael Feindt | 01h 12m |
History of work at CERN and the NeuroBayes algorithm from Blue Yonder ground claims in concrete results: lower out‑of‑stock rates at major supermarkets… | ||
| Lightning Talks Wednesday | Rafael Michelle | 54m |
Recorded at PyCon.DE 2017 Karlsruhe and edited with Blender Avidemux & Sonic Pi, the demo shows a working pipeline that regenerates documentation screenshots… | ||
| Lightning Talks Thursday | 48m | |
A fast-paced sequence of short demos and status reports covering concrete outcomes not abstractions: pandas on 50 CSVs parses ten‑thousand‑line files and… | ||
| Lightning Talks Friday | 42m | |
A rapid series of concrete demos and outcomes: an 11,000‑line Fortran → NumPy port runs with a JIT backend named Boreum and claims up to a 500× GPU speed‑up on… | ||
| Python in Space - The N Body Problem | Daniel Jilg | 41m |
Shows an actual Python implementation of Barnes-Hut and compares it to brute force: naive O(n^2) loops versus an O(n log n) quadtree/octree, with concrete… | ||
| Deep Learning for Computer Vision | Alex Conway | 42m |
Convolutional nets are presented as practical tools not miracles: pretrained VGG/ResNet weights + fine‑tuning get fast, high‑accuracy results (96.3% on 3.5k… | ||
| Modern ETL-ing with Python and Airflow (and Spark) | Tamara Mendt | 27m |
Claims that Python plus Apache Airflow and Apache Spark replace GUI ETL tools for production pipelines and shows it working at HelloFresh: Airflow runs… | ||
| Connecting PyData to other Big Data Landscapes using Arrow and Parquet | Uwe L. Korn | 32m |
Shows how Apache Parquet and Apache Arrow actually interoperate: Parquet provides 16 kilobytes pages, row groups of ~100k–2M rows, column statistics and… | ||
| Data Science Project for Beginners | Natalie Speiser, Jens Beyer | 31m |
Consulting-focused practical report on how projects actually run: Python, pandas and scikit-learn drive quick exploration, Jupyter hosts demos, models must… | ||
| Master 2.5 GB of unstructured specification documents with ease | Andreas Schilling | 30m |
2.5 GB of messy .doc/.docx become structured JSON and a local browser UI that loads and full‑text searches everything in about 6.5 seconds. The pipeline uses… | ||
| Lift your Speed Limits with Cython | Sefan Behnel | 42m |
Shows Cython compiling ordinary Python into C, adding typed C doubles and buffer access to drop object overhead, disabling bounds checks and releasing the GIL… | ||
| And now to something ELSE: Real Time Data Processing @ billiger.de | Axel Arnold | 33m |
billiger.de processes about 65 million offers and recomputes roughly six to eight million product updates per day using Python workers, Redis priority queues… | ||
| Hacking the Python AST | Suhas SG | 36m |
Clones CPython from GitHub and edits its grammar and AST pipeline to prove changes take effect at runtime, using the stdlib ast module with ast.NodeVisitor and… | ||
| The BorgBackup Project | Thomas Waldmann | 30m |
BorgBackup is a deduplicating backup built in Python 3.4+, Cython and C that actually encrypts, verifies and checkpoints repositories; it uses content-defined… | ||
| Vim your Python, Python your Vim | Miroslav Šedivý | 30m |
Shows how to type many languages on a US keyboard using the system compose key with a 6,000-line compose table, remap CapsLock so a tap is Escape and a hold is… | ||
| Plugin ecosystems for Python web-applications | Raphael Michel | 27m |
Claims a concrete, working plugin architecture used in pretix for years: a lightweight Django signal layer, automatic discovery via setuptools entry points… | ||
| The Python Ecosystem for Data Science: A Guided Tour | Christian Staudt | 37m |
Survey of specific libraries and workflows: NumPy pushes loops into C and yields an order-of-magnitude speedup; pandas does labeled tabular joins so a one-line… | ||
| Rasa: open source conversational AI to build next generation chatbots | Joey Faulkner | 31m |
Rasa provides open source Python libraries that actually show how intent classification uses pretrained word vectors averaged, entity extraction uses… | ||
| Synthetic Data for Machine Learning Applications | Hendrik Niemeyer | 45m |
ROSEN uses FEM simulations plus unsupervised methods to generate labeled training examples for pipeline anomaly risk models and claims those synthetics improve… | ||
| Sport analysis with Python | Thuy Le | 23m |
Dataset contains player x y positions and device id sampled at 20 millisecond timestamps and yields about ≈600,000 records for a ~46 minute match. Python code… | ||
| The Mustache Movement | Heidi Thorpe | 29m |
Demonstrates a working GAN built from existing Python modules (TensorFlow/PyTorch and numpy/scikit‑learn) trained on 8,000 celebrity images to learn… | ||
| Theoretical physics with sympy | Florian Thole | 33m |
Introduces SymPy basics and then applies them to a concrete Heisenberg-style model of 2D layered magnets, deriving symbolic expressions, simplifying with CSE… | ||
| Platform Intrusion Detection with Deep Learning | Carsten Pohl | 44m |
A production platform classifies bad bots with a concrete pipeline: data ingested into Google BigQuery, Keras and TensorFlow models trained from… | ||
| Why Python Has Taken Over Finance | Yves J. Hilpisch | 33m |
Argues that Python wins because concise code actually produces results: a few lines of pandas vectorized backtest builds positions, computes returns and plots… | ||
| Simple Data Engineering in python 3.5+ with Bonobo | Romain Dorgueil | 32m |
A compact Python 3.5+ ETL library for small scale pipelines that wires plain functions and generators into a transformation graph and runs them in threads… | ||
| High-Performance Ingestion with Python and Swarm64DB | Sebastian Dreßler | 41m |
Hardware-accelerated PostgreSQL plugin Swarm64DB plus Python and async I/O aim to push real ingestion of ~40‑byte rows from synthetic ~900k/sec benchmarks into… | ||
| turbodbc: Turbocharged Database Access for Data Scientists | Michael König | 33m |
Module turbodbc connects via standard ODBC drivers to deliver column-oriented results as NumPy arrays and Apache Arrow tables, extending DB-API 2.0 with… | ||
| Effective Data Analysis with Pandas Indexes | Alexander CS Hendorf | 31m |
Demonstration shows how Pandas indexing (especially TimeSeriesIndex) changes workflows: timestamps become index objects allowing dot-year/month access, slicing… | ||
| Programming the Web of Things with Python and MicroPython | Hardy Erlinger | 46m |
Shows MicroPython running on an esp8266 dev board (≈3€ chip) with a USB serial REPL at 115200 baud, blinking LEDs and reading a temperature sensor every 5s… | ||
| Empowered by Python - A success story | Jens Nie, Peer Wagner | 30m |
Company-wide migration to Python delivers concrete results: prototypes become saleable MVPs, tooling moves from MATLAB to Anaconda/Conda packaging… | ||
| Keeping the Grip on Decoupled Code Using CLIs | Anne Matthies | 21m |
Lightweight CLIs act as the glue that actually wires decoupled pieces together: Babbel packages small utilities with conda, exposes commands via setuptools… | ||
| Building your own SDN with Debian Linux, Salt Stack and Python | Maximilian Wilhelm | 37m |
Claims a working low‑budget SDN: Debian Linux plus Salt Stack drives config and orchestration, glue is 1000 lines of Python (about 14 public functions) that… | ||
| No Compromise: Use Ansible properly or stick to your scripts | Björn Meier | 28m |
Shows concrete fixes: replace fragile shell/command tasks with a custom Ansible module written in Python that implements proper version checks, supports check… | ||
| Data Plumbing 101 - ETL Pipelines for Everyday Projects | Eberhard Hansis | 28m |
Claims a concrete setup that turns dozens of brittle scripts into a traceable pipeline, reducing a tangled “17 import scripts” workflow to managed tasks… | ||
| Verified fakes with OpenAPI | Lauris Jullien | 07m |
Claim: keep test doubles honest by validating mocks against an API contract. Implementation details: use OpenAPI to validate requests and responses, generate… | ||
| The eye of the Python, an eye tracking system. From zero to... | Samuel Muñoz Hidalgo | 40m |
Webcam-based eye tracking is built end-to-end: a Pygame score-and-play game generates ~3,000 labeled images then augmentation multiplies that to ~90,000… | ||
| Large-scale machine learning pipelines using Luigi,...n | Alexander Bauer | 32m |
Claims a runnable, production-grade pipeline: Luigi + PySpark + scikit-learn orchestrates ETL, feature extraction, training and evaluation for demand… | ||
| Time series feature extraction with tsfresh - “get rich or die.. | Nils Braun | 27m |
tsfresh is a Python library that extracts and selects time‑series features at scale it runs >60 extractors to produce >500 features per series then applies… | ||
| An introduction to PyMC3 | Adrian Seyboldt | 44m |
Concrete demos show PyMC3 building full Bayesian models, running HMC and variational inference in a notebook, and diagnosing failures with divergences and… | ||
| Technical Lessons Learned from Pythonic Refactoring | Yenny Cheung | 31m |
Advocates measurable refactoring: write integration tests before refactor, change design without changing behavior, then add unit tests. Replaces dict params… | ||
| From Java to Python: Migrating Search Functionality at billiger.de | Patrick Schemitz | 29m |
Migrating custom search logic from Solr to SolrCloud and re-implementing six SolrComponents in a Python service layer yields concrete wins: it serves a 65… | ||
| Graphql in the Python World | Nafiul Islam | 25m |
Claims GraphQL can replace REST for fewer round trips and append-only schemas and shows concrete Python tooling: graphene as the implementation, the GraphiQL… | ||
| The Snake in the Tar Pit: Complex Systems with Python | Stephan Erb | 41m |
Argues that accidental complexity, not Python itself, consumes most engineering time and shows concrete defenses: replace brittle HTTP calls with timeouts… | ||
| Getting Scikit-Learn To Run On Top Of Pandas | Ami Tavory | 29m |
Claims that Scikit-Learn estimators become natively Pandas-aware by lazy dynamic wrapping at import time using a meta-path plus Python's inspect, converting… | ||
| Data Science Best Practices : From Proof of Concepts to Production | Yasir Khan | 29m |
Claims that 87% of data science projects fail and that the bottleneck is organization and integration rather than modeling. Shows concrete pipeline choices… | ||
| Really Deep Neural Networks with PyTorch | David Dao | 41m |
Shows concrete claims and code not buzzwords demonstrates PyTorch's tape-based automatic differentiation and GPU tensors and compares numpy implementations to… | ||
| Migrating existing codebases to using type annotations | Stephan Jeansch | 38m |
Shows how a large legacy Python codebase (tens or hundreds of thousands of lines) is incrementally annotated, enforced per‑file via pre-commit hooks, and… | ||
| Observing your applications with Sentry and Prometheus | Patrick Mühlbauer | 23m |
Explains how Sentry and Prometheus are actually used to catch production failures early with concrete examples and code snippets: Sentry aggregates events and… | ||
| An Admin's Cornucopia - Python Is More Than Just A Better Bash | Christian Theune | 30m |
Shows real code from infrastructure: tiny scripts through a Pyramid + SQLAlchemy inventory app, mixing Python 2.7 and 3.x. Demonstrates atomic rename via a… | ||
| Flow Is in the Air: Best Practices with Apache Airflow | Dominik Benz | 48m |
Summarizes two plus years of production use and claims Airflow converts analytics prototypes into running pipelines using programmatic DAGs and a pluggable… | ||
| Python on bare metal Beginners tutorial with MicroPython | Christine Spindler | 40m |
Claims a full Python runtime on tiny chips and shows it by handing out MicroPython pyboards and walking through REPL and main.py examples that blink the four… | ||
| Integrating Jupyter Notebooks into your Infrastructure | Florian Rhiem | 33m |
A custom JupyterHub setup integrates Jupyter Notebook servers with existing LDAP and storage services and runs one isolated container per user using Docker… | ||
| Python with Apache OpenWhisk | Ansgar Schmidt | 30m |
OpenWhisk is an open source serverless platform used as an alternative to AWS Lambda and MS Functions and the demo shows how to install locally, use the wsk… | ||
| From 0 to Continuous Delivery in 30 minutes. | David Wölfe | 31m |
Sets up a Continuous Delivery pipeline in 30 minutes using conda, GitLab CI and Anaconda Cloud and shows the actual repo files: a conda recipe, a… | ||
| Python is Weird | Dave Halter | 27m |
Covers the tokenizer reading left-to-right so odd tokens like ++4; and 0jif.1else-2 become valid, shows how tokenization produces surprising floats and… | ||
| Automated testing with 400TB memory | ChristophHeer | 33m |
Claims that a CI system handles SAP HANA testing at scale using Python, with a fleet totaling 400TB main memory across many hosts and about 1,300 servers in… | ||
| Project Avatar - Telepresence Robotics with Nao and Kinect | Thomas Reifenberg | 38m |
A telepresence system built from a Nao robot, an Oculus Rift and a Kinect One streams stereo video and depth data over Ethernet/Wi‑Fi, maps Kinect skeletons to… | ||
| nothing matches | ||