Talk Python To Me - Python Conversations For Passionate Developers

Large language models and chat-based AIs are kind of mind blowing at the moment. Many of us are playing with them for working on code or just as a fun alternative to search. But others of us are building applications with AI at the core. And when doing that, the slightly unpredictable nature and probabilistic nature of LLMs make writing and testing Python code very tricky. Enter promptimize from Maxime Beauchemin and Preset. It's a framework for non-deterministic testing of LLMs inside our applications. Let's dive inside the AIs with Max. Links from the show Max on Twitter: @mistercrunch Promptimize: github.com Introducing Promptimize ("the blog post"): preset.io Preset: preset.io Apache Superset: Modern Data Exploration Platform episode: talkpython.fm ChatGPT: chat.openai.com LeMUR: assemblyai.com Microsoft Security Copilot: blogs.microsoft.com AutoGPT: github.com Midjourney: midjourney.com Midjourney generated pytest tips thumbnail: talkpython.fm Midjourney generated radio astronomy thumbnail: talkpython.

If you're looking for fun data sets for learning, for teaching, maybe a conference talk, or even if you're just really into them, sports offers up a continuous stream of rich data that many people can relate to. Yet, accessing that data can be tricky. Sometimes it's locked away in obscure file formats. Other times, the data exists but without a clear API to access it. On this episode, we talk about PySport - something of an awesome list of a wide range of libraries (mostly but not all Python) for accessing a wide variety of sports data from the NFL, NBA, F1, and more. We have Koen Vossen, maintainer of PySport to talk through some of the more popular projects. Links from the show Koen on Twitter: @mr_le_fox PySport on Twitter: @PySportOrg Calling R from Python: medium.com DuckDB: duckdb.org PySport Playground: playground.pysport.org NFLVerse: github.com NBA Stats: nba.com Sports Databases: opensource.pysport.org Data sets: opensource.pysport.org Visualizations: opensource.pysport.org I/O: opensource.pysport

The release of Pydantic 2.0, its partial rewrite in Rust, and its refactoring into Pydantic core and top-level Pydantic in Python is big news. In fact, the alpha of Pydantic 2 was just released. Of course, these changes will have potentially wide ranging (and positive!) effects on libraries that are built upon Pydantic such as FastAPI, Beanie, and others. That's why this chance I had to catch up with Samuel Colvin from Pydantic and Sebastián Ramírez from FastAPI together, live from PyCon 2023. It's a super fun and wide ranging interview I'm sure you'll enjoy. Plus, there is a bit of an easter egg in the middle. Links from the show Sebastián Ramírez: @tiangolo Samuel Colvin: @samuel_colvin FastAPI: fastapi.tiangolo.com Pydantic: pydantic.dev Pydantic V2 Pre Release: pydantic.dev Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Spo

At PyCon 2023, there was a section of the expo floor dedicated to new Python-based companies called Startup Row. I wanted to bring their stories and the experience of talking with these new startups to you. So in this episode, we'll talk with founders from these companies for 5 to 10 minutes each. Links from the show Ponder: ponder.io generally intelligent: generallyintelligent.com Wherobots: wherobots.ai Neptyne: neptyne.com Nixtla: nixtla.io Predibase: predibase.com Pynecone: pynecone.io Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Sponsors Sentry Error Monitoring, Code TALKPYTHON Talk Python Training

Did you make this year's PyCon event in the US? There was a lot of excitement this time around in Salt Lake City. In this episode I'll bring you a bunch of experiences we had this year. It starts where frequent guest Jay Miller turns the tables and interviews me at the Microsoft booth on the expo hall floor in front of a live audience. Then you'll hear from Mario Munoz, Nick Muoh, Chris Williams, Ray McLendon, and Sean Tibor about their time at the conference. Links from the show Jay Miller: @kjaymiller Mario Munoz: @pythonbynight@fosstodon.org Ray McLendon: linkedin.com Nick Muoh: @nicksspirit@fosstodon.org Sean Tibor: @smtibor@fosstodon.org Chris Williams: @mistwire@fosstodon.org Python Community News: youtube.com The Birth & Death of JavaScript: destroyallsoftware.com Talk Python episode with Rivers Cuomo: talkpython.fm Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkp

What if we distributed CPython, the runtime, in the same way we distributed Python packages - as prebuilt binary wheels that only need to be downloaded and unzipped to run? For starters, that would mean we could ship and deploy Python apps without worrying whether Python itself is available or up-to-date on the platform. Nathaniel Smith has just proposed a PEP to do just that, PEP 711. And we'll dive into that with him next. Links from the show Nathaniel: @njs@mastodon.social [announce] Pybi and Posy: discuss.python.org PEP 711: peps.python.org Py2App: readthedocs.io PyInstaller: pyinstaller.org py-spy: github.com Anthropic: anthropic.com Trio: github.com Trio on Talk Python: talkpython.fm Zip Documentary: The Dark History of Zip Files: youtube.com Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Sponsors Sentry Error Monitoring,

What advice would you give someone just getting into Python? What did you learn over time through hard work and a few tears that would have really helped you? It's a fun game to play and we have Jason McDonald on the podcast to give us his take. Enjoy! Links from the show Jason C. McDonald: @codemouse92@mastodon.online Dead Simple Python: nostarch.com Coroutines and Tasks: docs.python.org Duck Typing: wikipedia.org Static Duck Typing in Python with Protocols: daan.fyi PEP 709: peps.python.org PEP 289: peps.python.org Python Packaging Strategy Discussion - Part 1: discuss.python.org Branch-detective: github.com Hypothesis: readthedocs.io Pydantic v2 announcement: pydantic.dev Michael's venv alias: digitaloceanspaces.com Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Sponsors Sentry Error Monitoring, Code TALKPYTHON Talk Python Tr

AI has taken the world by storm. It's gone from near zero to amazing in just a few years. We have ChatGPT, we have Stable Diffusion. But what about Jupyter Notebooks and pandas? In this episode, we meet Justin Waugh, the creator of Sketch. Sketch adds the ability to have conversational AI interactions about your pandas data frames (code and data). It's pretty powerful and I know you'll enjoy the conversation. Links from the show Sketch: github.com Lambdapromp: github.com Python Bytes 320 - Coverage of Sketch: pythonbytes.fm ChatGPT: chat.openai.com Midjourney: midjourney.com Github Copilot: github.com GitHub Copilot Litigation site: githubcopilotlitigation.com Attention is All You Need paper: research.google.com Live Colab Demo: colab.research.google.com AI Panda from Midjourney: digitaloceanspaces.com Ray: pypi.org Apache Arrow: arrow.apache.org Python Web Apps that Fly with CDNs Course: talkpython.fm Watch this episode on YouTube: youtube.com --- Stay in touch with us --- Subscribe to us on YouTube: you

We all know that privacy regulations are getting more strict. And that many of our users no longer believe that "privacy is dead". But for even medium-sized organizations, actually tracking how we are using personal info in our myriad of applications and services is very tricky and error prone. On this episode, we have Thomas La Piana from the Fides project to discuss privacy in our applications and how Fides can enforce and track privacy requirements in your Python apps. Links from the show California Consumer Privacy Act (CCPA): oag.ca.gov 30 Biggest GDPR Fines So Far: tessian.com Website fined for Google Fonts: theregister.com Fides on Github: github.com Fides: ethyca.com Bunny.net Fonts: fonts.bunny.net DBT: getdbt.com eBFP Kernel tools: ebpf.io nox: nox.thea.codes rich-click: github.com Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon:

In recent years, there has been a lot of experimenting how we work with dependencies and external libraries for our Python code. There is pip, pip-tools, Poetry, pdm, pyenv, pipenv, Hatch and others workflows. We dove into this deeply back on episode 406: Reimagining Python's Packaging Workflows. We're back with Ofek Lev to take a deeper look at Hatch. Links from the show Hatch: hatch.pypa.io Ofek on Twitter: @Ofekmeister Mamba: github.com Hatch env management: hatch.pypa.io Packaging a Python project tutorial: packaging.python.org Customize project generation: hatch.pypa.io Textual: textualize.io Ruff on Talk Python: talkpython.fm RustUp: rustup.rs Conda: docs.conda.io import antigravity: xkcd.com Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Sponsors Microsoft Founders Hub 2023 Sentry Error Monitoring, Code TALKPYTHON Talk P

If you're like most people, the simplicity and easy of getting started is a big part of pytest's appeal. But beneath that simplicity, there is a lot of power and depth. We have Brian Okken on this episode to dive into his latest pytest tips and tricks for beginners and power users. Links from the show pytest tips and tricks article: pythontest.com Getting started with pytest Course: training.talkpython.fm pytest book: pythontest.com Python Bytes podcast: pythonbytes.fm Brian on Mastodon: @brianokken@fosstodon.org Hypothesis: readthedocs.io Hypothesis: Reproducability: readthedocs.io Get More Done with the DRY Principle: zapier.com "The Key" Keyboard: stackoverflow.blog pytest plugins: docs.pytest.org Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Sponsors Microsoft Founders Hub 2023 Brilliant 2023 Talk Python Training

The great power of Python is its over 400,000 packages on PyPI to serve as building blocks for your app. How do you get those needed packages on to your dev machine and managed within your project? What about production and QA servers? I don't even know where to start if you're shipping built software to non-dev end users. There are many variations on how this works today. And where we should go from here has become a hot topic of discussion. So today, that's the topic for Talk Python. I have a great panel of guests: Steve Dower, Pradyun Gedam, Ofek Lev, and Paul Moore. Links from the show Python Packaging Strategy Discussion - Part 1: discuss.python.org Thoughts on the Python packaging ecosystem: pradyunsg.me Python Packaging Authority: pypa.io Hatch: hatch.pypa.io Pyscript: pyscript.net Dark Matter Developers: The Unseen 99%: hanselman.com Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python

So you know about dependencies and testing, right? If you're talking to a DB in your app, you have to decide how to approach that with your tests. There are lots of solid options you might pick and they vary by goals. Do you mock out the DB layer for isolation or do you use a test DB to make it as real as possible? Do you just punt and use the real DB for expediency? What if your dependency was a huge array of radio telescopes and a rack of hundreds of bespoke servers? That's the challenge on deck today were we discuss testing radio astronomy with pytest with our guest James Smith. He's a Digital Signal Processing engineer at the South African Radio Astronomy Observatory and has some great stories and tips to share. Links from the show GPU-based correlator for MeerKAT: github.com Meerkat: sarao.ac.za SARAO: sarao.ac.za Skarab server: peralex.com pycuda: documen.tician.de Commercial Telescopes: telescope.com PyLaTeX: github.com Linearity Test Code: talkpython.fm Correlator Context: talkpython.fm Watch this

Clean code is one of those aspects of your programming career that's easy to put on the back burner (sometimes by management more than yourself). But it's important in the short term for writing more debuggable and readable code. And important in the long run for avoiding having your program take on the dreaded "legacy code" moniker. We're fortunate to have Bob Belderbos back on the show. He's been thinking and writing about clean code and Python a lot lately and we'll dive into a bunch of tips you can use right away to make your code cleaner. Links from the show Bob on Mastodon: @bbelderbos@fosstodon.org PyBites: pybit.es Tips for clean code in Python article: pybit.es Refactoring book: pybitesbooks.com Final type: docs.python.org Sentinels pattern: python-patterns.guide Black formater: pypi.org Guarding clauses: medium.com ChatGPT: chat.openai.com Git Precommit: pre-commit.com #100DaysOfCode in Python course: training.talkpython.fm #100DaysOfWeb in Python course: training.talkpython.fm Watch this episode

Imagine a world with free and unlimited clean energy. That's the musings of a great science fiction story. But nuclear fusion (the kind that powers the sun) has always been close at hand, we see the sun every day, and yet impossibly far away as a technology. We took a major step towards this becoming a reality with the folks at the Lawrence Livermore National Labratory in the US achieved "ignition" where they got significantly more energy out than they put in. And Python played a major role in this research and experiment. We have Jay Salmonson here to give us a look at the science and the Python code of this discovery. Links from the show Jay on Mastodon: hachyderm.io/@jdsalmonson Jay on Twitter: @JaySalmonson Official Announcement: lasers.llnl.gov QnD Package: github.com PlasmaPy: plasmapy.org ML in Fusion: llnl.gov National Ignition Facility Achieves Ignition in Historic Nuclear Fusion Experiment: newenergytimes.net Video demonstrating the fusion lab: youtube.com Watch this episode on YouTube: youtube

When you think about processing tabular data in Python, what library comes to mind? Pandas, I'd guess. But there are other libraries out there and Polars is one of the more exciting new ones. It's built in Rust, embraces parallelism, and can be 10-20x faster than Pandas out of the box. We have Polars' creator, Ritchie Vink here to give us a look at this exciting new data frame library. Links from the show Ritchie on Mastodon: @ritchie46@fosstodon.org Ritchie on Twitter: @RitchieVink Ritchie's website: ritchievink.com Polars: pola.rs Apache Arrow: arrow.apache.org Polars Benchmarks: pola.rs Coming from Pandas Guide: github.io Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Sponsors Taipy User Interviews Talk Python Training

At some point, you've probably migrated an app from one framework or major runtime version to another. For example, Django to Flask, Python 2 to Python 3, or even Angular to Vue.js. This can be a big challenge. If you had 100s of active devs and millions of lines of code, it's a huge challenge. We have Ben Bariteau from Yelp here to recount their story moving 3.8M lines of code from Python 2 to Python 3. But this is not just a 2-to-3 story. It has many lessons on how to migrate code in many situations. There are plenty of gems to take from his experience. Links from the show Ben on Twitter: @benbariteau Ben's Talk at PyCon 2022: youtube.com python-modernize: github.com python-future: github.com Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Follow Michael on Mastodon: mkennedy Sponsors Cox Automotive User Interviews Talk Python Training

Our code quality tools (linters, test frameworks, and others) play an important role in keeping our code error free and conforming to the rules our teams have chosen. But when these tools become sluggish and slow down development, we often avoid running them or even turn them off. On this episode, we have Charlie Marsh here to introduce Ruff, a fast Python linter, written in Rust. To give you a sense of what he means with fast, common Python linters can take 30-60 seconds to lint the CPython codebase. Ruff takes 300 milliseconds. I ran it on the 20,000 lines of Python code for our courses web app at Talk Python Training, and it was instantaneous. It's the kind of tool that can change how you work. I hope you're excited to learn more about it. Links from the show Charlie on Twitter: @charliermarsh Charlie on Mastodon: @charliermarsh@hachyderm Ruff: github.com PyCharm Developer Advocate Job: jetbrains.com/careers Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch

Monorepos are contrary to how many of us have been taught to use source control. To start a project or app, the first thing we do is create a git repo for it. This leads to many focused and small repositories. A quick check of my GitHub account shows there are 179 non-fork repositories. That's a lot but I think many of us work that way. But it's not like this with monorepos. There you create one (or a couple) repositories for your entire company. This might have 100s or 1,000s of employees working on multiple projects within the single repo. Famously, Google, Meta, Microsoft, and Airbnb all employ very large monorepos with varying strategies of coordination. On this episode, we have David Vujic here to give us his perspective on monorepos as well as highlight an architectural pattern and set of tools for accomplishing this in Python. Links from the show David on Twitter: @davidvujic David on Mastodon: @davidvujic@mastodon.nu Monorepo definition: wikipedia.org git-sizer tool for large repos: github.

The iconic and first ever image of a black hole was recently released. It took over a decade of work and is a major achievement for astronomy and broadens our understanding of the universe for all of us. Would it surprise you to know that Python played a major part in this discovery? Of course it did, and Dr. Sara Issaoun is here to give us the full story. Links from the show Sara's PyCon keynote: youtube.com Sara on Twitter: @saraissaoun Event Horizon Telescope: eventhorizontelescope.org Black Hole Image Makes History; NASA Telescopes Coordinated Observations: nasa.gov Event Horizon Data: eventhorizontelescope.org Imaging, analysis, and simulation software for radio interferometry Package: github.com Initial data showing ring (matplotlib) (video at time): youtube.com Mars 2020 Helicopter GitHub Badge: github.blog Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to us on YouTube: youtube.com Follow Talk Python on Mastodon: talkpython Foll