promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_promptdojo_

Dimensionality and distance traps — step 7 of 7

ch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommendersch 40 · unsupervised learning, embeddings, and recommenders7/7
Checkpoint

One last thing before we move on. pass this to mark the lesson done, or skip and keep moving. hop to the next when you're ready.

Measure the curse of dimensionality yourself. Write contrast(d, n=60): seed already set — drop n random d-dimensional points and one query (all coordinates random.random()), then return nearest distance / farthest distance from the query. Print the ratio for dims=2 and dims=500 (exact format: f"dims=2 nearest/farthest = {low:.2f}" / f"dims=500 nearest/farthest = {high:.2f}"), then print whether contrast shrank (high > low + 0.3).

full-screen editor opens — close anytime to keep reading.

open the editor to run it.