Examples gallery

Tutorials and runnable examples

A scikit-learn-style example index: each page starts from a concrete question, shows the runnable code, and then shows the result.

Tutorial map

This page is a real example gallery built from frozen snapshots of public data. Every entry below links to a dedicated page generated from an actual EchoTime run, with visual outputs, runnable source code, local CSV snapshots, and links to reports or story assets where they exist.

Beginner examples

Short, low-friction walkthroughs that make the package legible in a few minutes.

Beginner example

US Treasury 10Y and 2Y yields in 2024

A first real-data example: read two columns from a CSV snapshot, compare them, and inspect how stable the relationship stays across the year.

Input two numeric columns from one CSVResult Pearson 0.82 / Spearman 0.77 / Mutual info 0.45

Open example View source

Beginner example

Python and JavaScript pageviews move together, but not identically

A clean two-column workflow on real Wikimedia traffic data: read a table, compare two series, then inspect how the relationship changes over rolling windows.

Input two numeric columns from one tableResult Pearson 0.75 / Spearman 0.86 / Mutual info 0.43

Open example Open report View source

Cross-disciplinary examples

Examples where timestamps, cohorts, or domain structure matter enough to change the similarity story.

Cross-disciplinary example

Irregular earthquake streams can still be compared

A real irregular-data example using 2024 USGS earthquakes from California and Alaska, with explicit timestamps kept in the comparison instead of being flattened away.

Input irregular values plus explicit timestampsResult Pearson 0.12 / Spearman 0.08 / Mutual info 0.07

Open example Open report View source

Flagship demos

Examples built to travel in talks, README sections, social posts, and benchmark discussions.

Flagship demo

AI attention breakout analogs

A flagship real-data example asking which historical attention curve DeepSeek looked more like over its first breakout window: ChatGPT or Threads.

Input one target series plus one or more analog curvesResult Pearson 0.94 / Spearman 1.00 / Mutual info 0.64

Open example Open report View source

Flagship demo

BTC, Brent, and VIX in 2024

A macro example on real FRED data asking whether BTC behaved more like oil or more like implied volatility across 2024.

Input two or more aligned market seriesResult Pearson 0.06 / Spearman 0.09 / Mutual info 0.12

Open example Open report View source

Flagship demo

Southwest heatwave city temperatures

A wide-table example using Phoenix and Las Vegas daily maximum temperatures during the 2024 summer heatwave window.

Input wide table or multichannel array with timestampsResult Pearson 0.73 / Spearman 0.75 / Mutual info 0.28

Open example Open report View source

Built-in starter datasets

The examples above use real public data. The table below is different: these are the small built-in starter datasets that ship with EchoTime for offline smoke tests and quick local experiments.

Dataset	Domain	Kind	Why it exists
Sine vs noise sine_vs_noise	generic	comparison	The smallest possible demo for rhythmicity and predictability.
Weekly website traffic weekly_website_traffic	traffic	dataset	A product-style dataset with weekly rhythm, trend, and launch bursts.
Irregular patient vitals irregular_patient_vitals	clinical	dataset	A minimal irregularly sampled clinical-monitoring cohort.
GitHub breakout analogs github_breakout_analogs	product	comparison	A social-friendly stars-over-time demo for similarity analysis.
BTC / gold / oil under shocks btc_gold_oil_shocks	markets	comparison	A regime-aware market comparison demo.
Grid load vs heatwave energy_load_heatwave	energy	dataset	A small multivariate energy system demo with heatwave-driven shifts.
Wearable recovery cohort wearable_recovery_cohort	wearable	dataset	A longitudinal cohort demo for heterogeneity and adherence.