Lightnews — Scholar-powered news

Michael Chavinda @mschav.bsky.social · 2d

Wrote a new article where I checkpoint the work we’ve done so far enabling Kaggle style EDA-to-model workflows in Haskell.

Covers Jupyter notebooks, dataframes, charting and machine learning with the iris dataset as the working example.

mchav.github.io/iris-classif...

Progress towards Kaggle-style workflows in Haskell

There’s been a lot of work in the Haskell ecosystem that has made it easier to write interactive Kaggle-like scripts. I’d like to showcase the synergy between 3 such tools: dataframe (my own creation)...

mchav.github.io

1 1 8

Michael Chavinda @mschav.bsky.social · 6d

It’s telling that we use this as meaningless filler text for the internet - a system driven almost entirely by the pursuit and marketing of pleasures.

2

Michael Chavinda @mschav.bsky.social · 6d

Lorem ipsum is taken from an excerpt about stoicism which says:

“The wise man, therefore, always holds in these matters to this principle of selection: he rejects pleasures to secure other greater pleasures, or else he endures pains to avoid worse pains.”

1 1 1

Reposted by Michael Chavinda

Polars @pola.rs · 6d

Swiss insurer La Mobilière refactored their risk model to Polars, achieving 5-10x speedups and enabling actuaries to run millions of simulation years on laptops. A scale previously unfeasible with pandas due to memory and single-core limitations.

pola.rs/posts/case-m...

Polars helps coping with black swan events at La Mobilière

DataFrames for the new era

pola.rs

1 7

Michael Chavinda @mschav.bsky.social · 11d

The MIT program synthesis class finally has problem sets. I’ll dedicate most of the fall/winter to completing it.

people.csail.mit.edu/asolar/Synth...

Introduction to Program Synthesis

people.csail.mit.edu

1

Michael Chavinda @mschav.bsky.social · 12d

On the bright side the things I’m good at are things I’ve inadvertently dedicated a lot of time to. Maybe the trick is just to be more aware of how my time is spent.

1

Michael Chavinda @mschav.bsky.social · 12d

There are a lot of things I wish I had the discipline to learn over the years. I convinced myself I needed the right mentor or book. Years later, more has been lost from indecision than wrong decision.

1 2

Michael Chavinda @mschav.bsky.social · 14d

Thanks for that context. @nickseagull.bsky.social i didn’t know you were basically a serial entrepreneur.

1

Michael Chavinda @mschav.bsky.social · 15d

Rebooting the work started by @ocramz.bsky.social with dataHaskell

Come through: discord.gg/UXcv5Eaz

Join the DataHaskell Discord Server!

Check out the DataHaskell community on Discord - hang out with 5 other members and enjoy free voice and text chat.

discord.gg

2 1 11

Michael Chavinda @mschav.bsky.social · 20d

This is such a subtle bug. In code this dense I tend to try and be as explicit as possible about my structures and types. Maybe the wrong takeaway but it would have been less of a headache to either: show explicitly what the two tuples components are and/or have a separate function this this.

1

Reposted by Michael Chavinda

The Urbanist @theurbanist.org · 26d

“When a full-service grocer leaves a neighborhood, it doesn’t just take away a shopping option; it reshapes the local food landscape. Competition shrinks, which means prices rise and choices narrow. Families end up paying more whether they stay local or spend extra time and money traveling farther.”

The Urbanist @theurbanist.org · 29d

NEW // Op-Ed: How to Confront Fred Meyer Closures and Guarantee Food Access

By Davida Sharpe

via @theurbanist.org // 🔗 www.theurbanist.org/2025/09/16/o...

Op-Ed: How to Confront Fred Meyer Closures and Guarantee Food Access » The Urbanist

# Kroger blames theft for four local grocery store closures, but the real story is more complicated. Kroger seeks to squeeze out more profit by driving up food insecurity. Standing up public or nonpro...

www.theurbanist.org

1 6 28

Michael Chavinda @mschav.bsky.social · 26d

Everything is converging to the same syntax.

Quansight @quansight.com · 27d

🐼 pandas continues to advance!

The new pd.col() syntax in pandas 3.0 makes data transformations cleaner and more predictable than lambda functions.

Marco Gorelli shows how expressions solve scoping issues while enabling more efficient method chaining.

💡 Discover more:

Expressions are coming to pandas!

`pd.col` will soon be a real thing!

buff.ly

1

Michael Chavinda @mschav.bsky.social · 28d

Doing important things with AI is misguided. We use tasks to build context. That context begets creativity and problem solving in the future. Doing stuff is both a creational and transformative process. AI is faster but using it undercuts the most important investment you could make in your task.

2

Michael Chavinda @mschav.bsky.social · Sep 15

Samthing Soweto to start the morning

youtu.be/42mq448-arg?...

Come Duze

YouTube video by Samthing Soweto - Topic

youtu.be

Michael Chavinda @mschav.bsky.social · Sep 13

Have some pretty cool examples of feature engineering using program synthesis on Haskell data frames.

Give a function space, we run a breadth first search to find what functions (and their compositions) have the highest correlation with a target variable.

github.com/mchav/datafr...

dataframe/app/Main.hs at feature_engineering · mchav/dataframe

A fast, safe, and intuitive DataFrame library. Contribute to mchav/dataframe development by creating an account on GitHub.

github.com

2

Reposted by Michael Chavinda

Haskell programming language @haskell.org · Sep 13

#AmeriHac – The hackathon by the #Haskell Foundation in New York City has been announced for the 7th and 8th of February 2026!

discourse.haskell.org/t/the-inaugu...

The Inaugural North America Haskell Hackathon

North American Hackathon The Haskell Foundation is proud to announce the inaugural AmeriHac, a two day haskell hackathon, with this iteration being in New York City! Jane Street has kindly offered to ...

discourse.haskell.org

3 6

Michael Chavinda @mschav.bsky.social · Sep 12

Starting a series on program synthesis:

mchav.github.io/an-introduct...

An introduction to program synthesis

Introduction This post kicks off a hands-on series about program synthesis—the art of generating small programs from examples. We’ll build a tiny, FlashFill-style synthesiser that learns to turn strin...

mchav.github.io

1

Michael Chavinda @mschav.bsky.social · Sep 9

Starting my foray into program synthesis and data wrangling.

Currently trying to implement some interesting papers.

1

Reposted by Michael Chavinda

Bryce @brycemecum.com · Sep 7

Check out @andrewlamb1111.bsky.social 's talk at the recent Iceberg meetup for a condensed overview of the the new Variant type coming to Parquet #apacheparquet #apachearrow