Skip to content

Picking good non-conformity scores for spatio-temporal and other data

License

Notifications You must be signed in to change notification settings

nandantumu/conformal_region_designer

Repository files navigation

Conformal Region Designer

Welcome to the Conformal Region Designer package! This package provides a set of tools and utilities for designing and visualizing conformal regions.

Example Image

Installation

To install the Conformal Region Designer package, you can use pip:

pip install conformal-region-designer

Usage

The following code creates a conformal region creator with a target of 90% coverage, using a kernel density estimator to estimate the density of the data, a mean shift algorithm to estimate the center of the data, and an ellipse to define the region.

import conformal_region_designer as crd
pcr = crd.ConformalRegion('kde', 'meanshift', 'ellipse', 0.90)
pcr.fit(Z_cal_one)
pcr.conformalize(Z_cal_one)

The regions can be visualized using the following code:

fig, ax = plt.subplots()
pcr.plot(ax=ax)

About

Picking good non-conformity scores for spatio-temporal and other data

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published