Meta tags:
description=A collection of my thoughts on the various topics I find myself interested in.;
author=Michael Chavinda;

Headings (most frequently used words):

dataframes, for, introduction, what, an, haskell, program, synthesis, is, about, and, programming, learning, to, michael, chavinda, context, complexity, the, big, of, agent, api, first, look, at, token, efficiency, pandas, feels, clunky, coming, from, grow, mow, interpretable, models, with, boosting, symbolic, regression, graphs, type, level, still, category, theory, teaches, us, better, decision, tree, splits, llms, as, heuristics, installing, docker, on, chromebook, part, ii, automatically, generating, features, machine, progress, towards, kaggle, style, workflows, in, rewriting, microhs, benchmarking, against, python, my, fondness, alternative, haskells,

Text of the page (most frequently used words):
the (34), more (16), read (13), and (11), for (10), about (10), that (10), you (10), from (8), what (7), haskell (6), #dataframe (5), post (5), program (5), synthesis (5), with (5), #pandas (5), been (4), has (4), dataframes (4), code (4), like (4), find (4), introduction (4), can (4), was (4), api (4), models (4), are (4), agent (4), performance (3), but (3), thought (3), this (3), easier (3), kaggle (3), tools (3), them (3), different (3), your (3), article (3), notebooks (3), working (2), year (2), while (2), come (2), haven (2), made (2), python (2), kicks (2), off (2), hands (2), series (2), how (2), tiny (2), nkomo (2), into (2), see (2), write (2), search (2), one (2), problem (2), style (2), lot (2), part (2), apply (2), automatically (2), generating (2), features (2), dataset (2), learning (2), through (2), something (2), name (2), two (2), columns (2), simple (2), messy (2), some (2), seem (2), apis (2), look (2), data (2), type (2), programming (2), have (2), use (2), model (2), complexity (2), interpretable (2), ago (2), blog (2), side (2), author (2), than (2), think (2), when (2), making (2), tool (2), its (2), feels (2), clunky (2), coming (2), since (2), small (2), great (2), token (2), good (2), collection (2), thoughts (2), various (2), topics (2), myself (2), interested (2), michael (2), chavinda (2), implementation, now, focus, ergonomics, question, inevitably, significant, investments, might, worth, snapshotting, establish, baseline, benchmarking, against, fondness, alternative, haskells, rewriting, microhs, art, teaching, machines, generate, build, synthesiser, learns, turn, strings, joshua, input, output, pairs, define, string, manipulation, language, interpreter, space, programs, solves, our, toy, flashfill, there, work, ecosystem, interactive, scripts, showcase, synergy, between, such, own, creation, ihaskell, hasktorch, progress, towards, workflows, second, previously, explored, technique, enumerative, bottom, slightly, realistic, iris, machine, couldn, any, instructions, online, here, anyone, who, goes, similar, struggle, installing, docker, chromebook, tabular, modeling, gets, moment, stumble, onto, right, derived, quantity, not, mysterious, deep, usually, ratio, turns, raw, rate, difference, becomes, margin, count, captures, bunch, fields, were, hinting, better, decision, tree, splits, llms, heuristics, every, library, ships, hundreds, operations, alone, over, 200, methods, these, same, operation, wearing, hats, others, genuinely, distinct, without, framework, telling, apart, end, memorizing, instead, understanding, structure, pipe, applymap, agg, transform, map, melt, pivot, category, theory, teaches, showing, friend, typed, whole, pitch, derive, schema, then, compiler, catches, column, typos, mismatches, all, stuff, would, otherwise, blow, runtime, had, nice, demo, ready, using, 284, 000, rows, credit, card, fraud, level, still, convergence, ideas, floating, head, learn, stochastic, algorithmic, algebraic, rein, make, grow, mow, boosting, symbolic, regression, graphs, years, came, across, issue, mentioned, titled, showed, exploration, compared, time, concluded, clunkier, operationalises, definition, clunkiness, really, know, thing, feel, drift, further, away, task, idiosyncracies, why, frames, repo, saw, largely, discredited, didn, clear, methodology, unfortunately, yet, anecdotal, experiment, comparing, generation, sabela, marimo, environment, agents, they, modular, design, create, top, also, much, intervene, which, languages, most, efficient, first, efficiency, obvious, repair, cheaper, done, elicited, another, retry, provided, discloses, changed, only, outputs, specificity, parsimony, truthfulness, local, instrument, measuring, harness, costs, tldr, context, big,

Text of the page (random words):
michael chavinda a collection of my thoughts on the various topics i find myself interested in michael chavinda a collection of my thoughts on the various topics i find myself interested in blog about context complexity what is the big o of an agent api tldr small local models are a good instrument for measuring what a harness costs an agent an api is only as good as its outputs specificity parsimony and truthfulness obvious code repair is cheaper done in your tools than elicited from the model through another retry provided the tool discloses what it changed read more a first look at token efficiency a while ago i saw the article which languages are most token efficient the article was largely discredited since it didn t have a clear methodology unfortunately i haven t come up with one yet but i thought to do a small anecdotal experiment comparing code generation in haskell notebooks with sabela vs python notebooks marimo notebooks are a great environment to use with agents since they are modular by design and you can create great agent apis on top of them it s also much easier to intervene as the agent is working read more pandas feels clunky coming from r what about haskell some years ago i came across an issue in the frames repo that mentioned a blog post titled why pandas feels clunky when coming from r the article showed a side by side of simple data exploration in r and compared the code to pandas at the time the author concluded that pandas was clunkier than r the author operationalises the definition of clunkiness but i think it s really more of a you know it when you see it thing you can feel if an api is making you drift further away from your task and making you think more about the tool and its idiosyncracies read more grow and mow interpretable models with boosting symbolic regression and e graphs this post is the convergence of two ideas that have been floating in my head for about a year can we learn messy stochastic models and use algorithmic algebraic tools to rein in model complexity to make models interpretable read more type level programming is still programming i was showing a friend the typed dataframe api the whole pitch was look you derive a schema from your data and then the compiler catches column name typos type mismatches all the stuff that would otherwise blow up at runtime i had a nice demo ready using the kaggle credit card fraud dataset about 284 000 rows 31 columns read more what category theory teaches us about dataframes every dataframe library ships with hundreds of operations pandas alone has over 200 methods on a dataframe is pivot different from melt is apply different from map what about transform agg applymap pipe some of these seem like the same operation wearing different hats others seem genuinely distinct without a framework for telling them apart you end up memorizing apis instead of understanding structure read more learning better decision tree splits llms as heuristics for program synthesis a lot of tabular modeling gets easier the moment you stumble onto the right derived quantity not something mysterious or deep it s usually something you can name a ratio that turns two raw columns into a rate a difference that becomes a margin a simple count that captures what a bunch of messy fields were hinting at read more installing docker on a chromebook i couldn t find any instructions online so i thought i d post them here for anyone who goes through a similar struggle read more an introduction to program synthesis part ii automatically generating features for machine learning introduction this post kicks off the second part of a hands on series about program synthesis we ll apply the previously explored technique an enumerative bottom up search to a slightly more realistic problem automatically generating features for the iris dataset read more 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 hasktorch and ihaskell read more an introduction to program synthesis introduction this post kicks off a hands on series about program synthesis the art of teaching machines how to generate code we ll build a tiny flashfill style synthesiser that learns to turn strings like joshua nkomo into j nkomo from input output pairs we ll see how to define a tiny string manipulation language write an interpreter and search the space of programs to find one that solves our toy problem read more rewriting dataframes for microhs my fondness for alternative haskells read more benchmarking haskell dataframes against python dataframes i ve been working on a dataframe implementation in haskell for about a year now while my focus has been on ergonomics the question of performance has inevitably come up i haven t made significant performance investments but i thought it might be worth snapshotting the performance to establish a baseline read more