A Python package for computing confidence and prediction intervals for 0D and 1D univariate and bivariate data.
This package contains:
- Functions and classes for computing and displaying confidence regions and predictions regions for 0D and 1D datasets
- Scripts for replicating results from the main paper (see below)
- Notebooks for reviewing theory and for standalone, step-by-step calculations
Requirements:
- Python 3.x
- NumPy
- SciPy
- Matplotlib
- spm1d
Please consider citing:
Pataky TC, Yamagata M, Ichihashi N, Duarte M (in review). Adjustment to: "CI2 for creating and comparing confidence-intervals for time-series bivariate plots" Gait & Posture.
Note: ci1d does not support ellipsoid or hyperellipsoids calculations for multivariate data, they are not implemented in this package because they cannot be readily visualized.