BuildMidData / MLAuto-graded

Split a dataset reproducibly with balanced classes

Protecting a fair evaluation from mistakes that never crash: duplicated, dropped, or unevenly sampled records.Every model evaluation rests on an honest split of the data into a training set and a held-out test set. Get the split wrong and every metric downstream is a lie. Implement split(X, y, test_size, seed, stratify=True) returning four lists with each row still matched to its label — reproducible from the seed, and preserving class proportions when stratification is asked for.

35 min

python

Start lab

You're done when

Every row appears exactly once across the two sets, with none lost or duplicated, even when the data contains identical rows.
The same seed reproduces the same split exactly.
With stratification on, each class appears in the test set in the expected proportion, with exact counts.
The requested test proportion is honoured.

What it measures

Machine learningUnderstanding why stratification and reproducibility are prerequisites for a trustworthy evaluation.

Data preparation and preprocessingPartitioning data without losing or duplicating rows, including when duplicate payloads make identity ambiguous.

ReproducibilitySeeding so a result can be reproduced exactly, which is what makes a reported metric checkable.

Your environment

Terminal + editorA full shell in the browser

Claude CodePre-authenticated, no key setup

Isolated microVMYours alone, wiped afterwards

Start this lab

35 min

Start labYou'll sign in to Anthropos to start. The lab opens in a private cloud workspace with Claude Code ready to use.

AI Labs is part of Anthropos Workforce. Not on your plan yet? Book a demo