PyData NYC 2018
2018-10-11..2018-10-12
New York, United States
[PyData NYC 2018, PyData NYC 2022, PyData NYC 2023, PyData NYC 2024]
| title | author | length |
|---|---|---|
| Why I Use Julia For Quantum Physics | Katharine Hyatt | 42m |
Claims Julia replaces the two‑language workflow by letting simulation, analysis and job submission live in one‑language solution; tensor‑network code uses… | ||
| Scaling Interactive Pandas Workflows with Modin | Devin Petersohn | 38m |
Modin is a drop-in middle layer for pandas that speeds up interactive workflows by changing one import import modin.pandas as pd and running on Ray with block… | ||
| The Value of Null Results | Angel D'az | 30m |
Collected a pizza‑delivery dataset of 1301 rows and moves from Excel to Python using Query Storm for SQL queries then pandas and plotting. Tests include a… | ||
| The TileDB Array Data Storage Manager | Stavros Papadopoulos, Jake Bolewski | 37m |
TileDB claims an open source C++ storage manager for multi-dimensional sparse and dense arrays that maps n‑D cells into one‑dimensional tiled files, stores… | ||
| The Lifecycle of Artificial Intelligence with IBM's Deep Learning as a Service | Justin McCoy | 38m |
Claims a cloud service that trains evaluates and deploys models purely via API calls using fiddle (Fabric for Deep Learning) and Watson Studio, packages… | ||
| Deploying Data Science for Distribution of The New York Times | Anne Bauer | 38m |
Built a deployed pipeline that replaces COBOL heuristics with Poisson regression with censoring-aware likelihood and an analytic newsvendor policy (augmented… | ||
| A Bluffer's Guide to Dimension Reduction | Leland McInnes | 37m |
The claim is that almost everything useful about dimension reduction reduces to two concrete mechanisms: matrix factorization — build a reconstruction loss… | ||
| End to End Data Science Without Leaving The GPU | Randy Zwitch | 36m |
Claims an end-to-end GPU workflow using JupyterLab, Ibis, the OmniSci Jupyter kernel and Apache Arrow zero-copy Arrow references so data never serializes off… | ||
| Data Science in Healthcare: Beyond the Hype | Michael Becker | 33m |
Slides catalog high-profile claims that later fail: IBM Watson touted as better at cancer diagnosis then found to give unsafe recommendations after a… | ||
| Contributing to Open Source: A Guide | Paul Ganssle | 32m |
Shows a concrete contribution pipeline: fork, branch, add tests, push and open a GitHub pull request that passes continuous integration, with emphasis on a… | ||
| Using Embeddings to Understand the Variance and Evolution of Data Science... | Maryam Jahanshahi | 39m |
Dynamic exponential family embeddings applied to a million-plus job descriptions over three years use gensim and TensorFlow with heavy custom preprocessing and… | ||
| Two Years of Bayesian Bandits for E Commerce | Austin Rochford | 33m |
Monetate deploys Bayesian bandits since early 2016 to optimize e‑commerce; implementation uses a beta‑binomial likelihood and Thompson sampling with Monte… | ||
| Contract Testing in Python for a Data Science API | J. Scott Hajek | 29m |
Demo of using JSON Schema as a formal contract for a Python-backed REST microservice, showing concrete tests that validate request and response shape rather… | ||
| The new kid on the block: deep learning with GluonNLP | Sneha Jha | 38m |
Demonstrates GluonNLP on MXNet with concrete utilities: a sorted bucketing reduces average padding 11.7→3.7 for WikiText-2, built-in pretrained models (a… | ||
| How to spend ¾ of your yearly budget in 3 weeks: a PySpark cautionary tale | Nicole Carlson | 27m |
A recommendation engine built with PySpark on Databricks reads orders then 1000x larger member pageviews from Snowflake and S3 and runs cosine similarity… | ||
| pandas: Integer NA as a first class citizen | Jeff Reback | 31m |
pandas gains a concrete claim: add Integer NA as a first-class dtype via Extension Arrays, implemented as data plus a boolean mask over numpy storage. It… | ||
| Graph Analytics - From the Whiteboard to Your Toolbox | Sam Lerma | 31m |
Claims graph framing gives practical wins and shows exactly how to model problems, what code looks like, and what breaks. Examples convert social… | ||
| You are not a Computer: Python, Dynamical Systems, and Radical Embodied... | Henry S. Harrison | 30m |
Argues the brain is better modeled as a coupled dynamical system than a Turing‑style computer and shows concrete Python implementations in the PyData stack… | ||
| Repeatable Data Setup for Repeatable Science using Julia | Sebastin Santy | 26m |
PyData NYC 2018
This talk presents a framework DataDeps.jl, and its accompanying tool DataDepsGenerators.jl, for working with datasets in a more automatic and… | ||
| Functional Programming for Data Scientists | Santiago Basulto | 37m |
Argues that functional programming yields concrete wins in data work: shows numpy and pandas behavior (immutability, inplace=False, vectorized ops)… | ||
| Iodide and Pyodide: Bringing Data Science Computation to the Web Browser | Michael Droettboom | 33m |
Iodide and Pyodide move real data science into the browser by compiling CPython compiled to WebAssembly and shipping numpy and pandas in the browser so… | ||
| Unit Testing for Data Scientists | Hanna Torrence | 39m |
The talk argues that unit tests give concrete confidence by isolating specific functionality and recommends starting small while refactoring. It demonstrates… | ||
| HDBSCAN, Fast Density Based Clustering, the How and the Why | John Healy | 34m |
HDBSCAN becomes a practical tool by converting the naive O(n^2) approach into an O(n log n) implementation using dual‑tree Borůvka and spatial indexing… | ||
| … | James Powell | 38m |
Shows that Python accepts surprising Unicode identifier characters by inspecting CPython tokens, the grammar, PEP 3131, Unicode DerivedCoreProperties and… | ||
| Run Numba functions in SQLite: WTF? | Phillip Cloud | 26m |
An experimental library called slumba embeds Numba-generated C-callable functions into SQLite UDFs, aggregates and window functions by emitting LLVM/JIT code… | ||
| A data scientist's guide to production web apps with Flask | Ali Vanderveld | 34m |
Practical checklist: precompute feeds in a MySQL RDS table rather than running models at request time when the parameter space is bounded; use Flask-SQLAlchemy… | ||
| Conda Forge - Community Driven Packaging That Works for You | Marius van Niekerk | 37m |
Conda-forge packages cross-platform binaries for Python and non-Python deps, runs CI bots that auto-open rebuild PRs, and aims for exact reproducibility via… | ||
| Introduction to Big Data Processing using Spark and Python | Raoul-Gabriel Urma | 01h 22m |
Hands-on workshop that walks from Hadoop concepts to running PySpark code: it explains batch vs streaming, shows RDD and DataFrame patterns (map, flatMap… | ||
| Up your Bus Number - A Reproducible Data Science Workflow | Kjell Wooding, Amy Wooding | 01h 14m |
Claim: a reproducible end-to-end data science workflow that makes the messy 80% of data munging explicit and recoverable. Tools actually used include… | ||
| Hierarchical Probabilistic Modelling in Real Life | Benjamin Batorsky, Matthew Moocarme, Ph.D. | 01h 29m |
Claims a working pipeline that uses PyMC3 and NUTS sampler to turn realized spend into budget ranges, testing three models (pooled, product-level… | ||
| Open the Black Box: an Introduction to Model Interpretability with LIME and SHAP | Kevin Lemagnen | 01h 36m |
Workshop demonstrates concrete claims and code: uses Eli5 for white‑box weights, LIME for local, model‑agnostic approximations, and SHAP (tree explainer) for… | ||
| Crunching Your Data with CatBoost - the New Gradient Boosting Library | Vasily Ershov | 01h 33m |
PyData NYC 2018
A comprehensive tutorial on CatBoost (http://catboost.ai), a new open-source gradient boosting library, that outperforms existing publicly… | ||
| Making Faces: Conditional Generation of Faces using GANs... | Sophia R Searcy, Justin Blinder | 01h 27m |
Claims a conditional GAN that flips binary face attributes like smiling or bangs using spectral normalization and a versatile auxiliary classifier, implemented… | ||
| Going Beyond “Sorry, I Didn’t Get That”: Building AI Assistants that Scale... | Justina Petraitytė | 01h 29m |
Practical workshop demonstrating how to build scalable ML-only conversational assistants using Rasa NLU and Rasa Core. Covers intent classification, entity… | ||
| Measuring Model Fairness | J. Henry Hinnefeld | 28m |
Introduces concrete fairness metrics—disparate impact (statistical parity), equal opportunity—and compares their tradeoffs with examples such as COMPAS error… | ||
| Intake: Taking the Pain Out of Data Access | Martin Durant | 37m |
Intake provides a single interface and YAML-backed catalogs that describe data sources (CSV, Avro, S3, Azure, HTTP) and return a pandas DataFrame (or… | ||
| Train, Evaluate, Repeat: Building a Credit Card Fraud Detection System | Leela Senthil Nathan | 39m |
Stripe faces three concrete problems and gives concrete fixes: pick global labels like chargebacks and card‑network flags such as TC40 rather than merchant… | ||
| Unit Testing Data with Marbles | Jane Stewart Adams, Leif Walsh | 44m |
Argues that data needs executable assumptions and shows Two Sigma’s open‑source Marbles extending Python unit tests to capture local variables, add semantic… | ||
| nothing matches | ||