Awesome

LCPexperiments

You can either install LCPexperiments as a package via install_github("LeyingGuan/LCPexperiments") or source the files in under the R directory to run numerical examples in [1].

Illustrating example in Intro

See example_intro.Rmd

Experiments with simulated synthetic data

X = 1D_setA, 1D_setB, 1D_setC, 1D_setD represents one of the four different data generative settings.

Run example1: Rscript synthetic_example1.R X Run example2: Rscript synthetic_example1.R X

The intermediate training model is saved under synthetic_results/example1/* and synthetic_results/example1/* , and the final results are saved under synthetic_results/* .

Experiments with real data from UCI databases

X = CASP, Concrete, facebook, facebook2 represents one of the four different data sets.

Run X: Rscript simulation_X.R

The intermediate training model is saved under UCI_results/X/* , and the final results are saved under UCI_results/* .

[1] Localized Conformal Prediction: A Generalized Inference Framework for Conformal Prediction https://arxiv.org/abs/2106.08460.