IID Got You Down? Resample Time Series Like A Pro
summary
tsbootstrap is an open source Python library that implements moving/circular/stationary/non‑overlapping block bootstrap plus residual and Markov‑style bootstraps, with options for geometric or custom block‑length distributions, tapering weights, and model‑based resampling. A 15‑point demo uses block length 3 to show blocks break trend while residual bootstrap preserves it. Code is modular, uses hypothesis for property testing and shows 85% test coverage, and installs with pip install tsbootstrap and plans integration with sktime.
AI-written summary · about
chapters (7)
- 0:00TS bootstrap intro
- 4:04Classical bootstrapping fails
- 8:41Block bootstrapping methods
- 15:26Residual bootstrapping trend
- 20:00TS bootstrap features
- 23:28Quick demo usage
- 25:00Q&A
related talks
16mIntro to Workboots: Make Prediction Intervals from Tidymodel Workflows- 16mIntro to Workboots: Make Prediction Intervals from Tidymodel Workflows
01h 26mSktime—Python Toolbox for Time Series: How to Implement Your Own Estimator
30msktime: python toolbox for time series- next-generation AI
26mTime Series Analysis in Python: Easy Tools for Scientific Insight
34mError Estimation for Randomized Numerical Linear Algebra: Bootstrap Methods
01h 27mIntroduction to Machine Learning with Time Series
27mTime series feature extraction with tsfresh - “get rich or die..