This repository contains scripts to run the simulation described in Nonparametric Statistical Inference via Metric Distribution Function in Metric Spaces.

Please install Ball 1.3.13 to use the tests mentioned in the paper.

Note: download this package from here. Some advanced features haven't been integrated into the Ball package (version 1.3.13) in R CRAN. We expect to make the Ball package on R CRAN support these new features in version 1.3.14.

We have organized our scripts to improve the reproducibility of experiments. Specifically, each R script file corresponds to a certain part of the results in the paper, which are listed below:

• large_homogeneity_n.R <-> Results in Figure 3A
• large_joint_independence_n.R <-> Results in Figure 3B
• compare_homo_methods.R <-> Results in Figure 4A
• compare_indep_methods.R <-> Results in Figure 4B
• real_data_adni_analysis.R <-> Results in Table 1
• real_data_adhd200_analysis.R <-> Results in Table 2

Notice that, before conducting these scripts, please modify the your_path in each script accordingly.

Please cite the following publications if you make use of the material here.

• Xueqin Wang, Jin Zhu, Wenliang Pan, Junzhu Zhu, and Heping Zhang (JASA, 2023+). Nonparametric statistical inference via metric distribution function in metric spaces. arXiv preprint arXiv:2107.07317.

The corresponding BibteX entries:

@article{wang2023nonparametric,
  title={Nonparametric statistical inference via metric distribution function in metric spaces},
  author={Wang, Xueqin and Zhu, Jin and Pan, Wenliang and Zhu, Junhao and Zhang, Heping},
  journal={Journal of American Statistical Association},
  year={2023+},
  doi={10.1080/01621459.2023.2277417},
}

Please direct questions and comments to the issues page.