# Supplementary Materials - NeurIPS Submission

This document provides detailed instructions for running the simulation code and reproducing the experimental results presented in the paper. The parameter `r` controls the level of privacy protection, and its conversion to the differential privacy parameter `ε` is provided in the appendix.

---

## 📦 Case A.R

**Description:**  
This R script simulates the regression setting for **Case A**. The privacy protection level can be adjusted by modifying the parameter `r`.

**Instructions:**  
- Ensure that the required R packages are installed (`parallel`, `MASS`).
- Set the save path in the script according to your local environment.
- Run the script to replicate the results reported in the paper.
- You can also modify the data generation process to test on custom datasets, including **Case B** and **Case C**.

---

## 🐍 Case 1.ipynb

**Description:**  
This Jupyter notebook implements the classification setting for **Case 1**. The parameter `r` adjusts the level of privacy protection.

**Instructions:**  
- Ensure that the required Python libraries are installed (`numpy`, `scipy`, `pandas`, `matplotlib`, `tqdm`).
- Modify the save path in the notebook to your local environment.
- Execute the notebook to reproduce the results in the paper.
- You may also alter the data generation method to evaluate other cases, such as **Case 2** and **Case 3**.

---

## ⚡ Elec2.R

**Description:**  
This R script reproduces the real data experiment using the **Elec2** dataset. The dataset `electricity-normalized.csv` is provided. Alternatively, you can obtain the dataset from external sources.

**Instructions:**  
- Ensure that the required R packages are installed (`data.table`, `forecast`).
- Adjust the file import path and save path in the script to match your local directory structure.
- Run the script to reproduce the real data results as reported in the paper.

---
##  WISDMuser10.R

**Description:**  
This R script reproduces the real data experiment using the **WISDM User 10** dataset. The required nonconformity score file `user10_evenline_nonconformity.csv` is provided. This file contains the precomputed non-conformity scores necessary for running the online conformal prediction evaluation.

**Instructions:**  
- Ensure that the required R packages are installed.
- Adjust the file import path for `user10_evenline_nonconformity.csv` and the save path for results in the script to match your local directory structure.
- Run the script to reproduce the classification results as reported in the appendix of the paper.

---

## ✅ Notes:
- Adjusting the parameter `r` modifies the privacy protection level. The conversion to the `ε` parameter is detailed in the appendix.
- Ensure that all necessary libraries are installed before running the scripts or notebook.
- Modify file paths as needed to align with your local environment.

For further details, please refer to the appendix or address them during the rebuttal phase.
