Data and scripts underlying the publication: Quantifying the Spatial Scales of Animal Clusters Using Density Surfaces

<strong>Supplementary scripts to the publication "Quantifying the spatial scales of animal clustering using Density Surfaces"</strong><br>We implement a method to quantify the degree of clustering of point location data at different spatial scales, which uses Kernel Density Estimation to construct a density function from the underlying point-location data. We build upon this method to automatically detect cluster diameters using smoothing kernels that better represent the perception neighbourhood of animals.More details can be found in the manuscript.<br>These scripts construct the artificial data sets and results in the figures in the main text of the manuscript. <br><strong>data_generator.py</strong>This file contains the functions to construct the artificial data sets, as well as visualization tools to plot the point sets.Running the main() function:1. constructs all artificial data sets2. creates visualizations of all generated and real-life datasets, saves them as .pdf files, and shows them on-screen<br><strong>metric_calculator.py</strong>This file contains the functions to calculate the metric described in the manuscript, as well as to compute Ripley's K function and the Radial Distribution Function.Running the main() function:1. generates the metric functions for all artificial and real-life data sets2. creates visualizations of all generated metric functions, saves them as .pdf files, and shows them on-screen3. prints the found relevant spatial scales, and their metric values, in the terminal<br><strong>elephant.pickle</strong>This file contains the real-world dataset of elephant locations to be used in metric_calculator.pyThe original data was collected in March 2014 in the Tsavo National Parks, Kenya. We use a subset of the original data set, consisting of location data of 24 elephants obtained from an aerial image that were manually taken by human observers upon spotting the animals. The aerial image was manually processed into spatial data by placing a point on the approximate centre point of each animal in the image, and projected onto a 100x100 xy-plane.The data is serialized and de-serialized using the native Python package "pickle". The data format used by pickle is Python-specific.<br>To perform the experiments:1. Ensure you have a functioning Python3 installation.2. Install the required packages using pip: - numpy - matplotlib - scipy - scikit-learn3. Run the main() function in data_generator.py to generate the artificial datasets4. Run the main() function in metric_calculator.py to generate the metric functions and figures<br>

**论文《使用密度曲面量化动物集群空间尺度》（Quantifying the spatial scales of animal clustering using Density Surfaces）配套补充脚本**
本研究实现了一种可量化不同空间尺度下点位数据集群程度的方法：该方法借助核密度估计（Kernel Density Estimation）从原始点位数据中构建密度函数，并在此基础上，采用更贴合动物感知邻域的平滑核函数，自动检测集群直径。更多细节可参阅论文原稿。
本套脚本可生成人工数据集，并复现论文正文内的所有图表结果。
**data_generator.py**
该文件包含构建人工数据集的相关函数，以及用于绘制点位集的可视化工具。运行其main()函数可实现以下操作：
1. 生成全部人工数据集；
2. 对所有生成的人工数据集与真实数据集进行可视化，将结果保存为PDF格式文件并在屏幕上展示。
**metric_calculator.py**
该文件包含计算论文所述指标的相关函数，同时可实现里普利K函数（Ripley's K function）与径向分布函数（Radial Distribution Function）的计算。运行其main()函数可实现以下操作：
1. 为全部人工与真实数据集生成指标函数；
2. 对所有生成的指标函数进行可视化，将结果保存为PDF格式文件并在屏幕上展示；
3. 在终端中输出检测到的关键空间尺度及其对应指标数值。
**elephant.pickle**
该文件包含用于metric_calculator.py的真实大象点位数据集。原始数据采集于2014年3月肯尼亚察沃国家公园，本研究使用原始数据集的子集：包含24头大象的位置数据，这些数据源自人类观测者通过航拍图像手动标记的点位。航拍图像经人工处理：在图像中每头动物的近似中心点处标记点位，并将其投影至100×100的xy平面。该数据使用Python原生序列化模块pickle进行序列化与反序列化，pickle的数据格式为Python专属格式。
**实验执行步骤**
1. 确保已配置可用的Python3运行环境；
2. 通过pip安装所需依赖包：
- numpy
- matplotlib
- scipy
- scikit-learn
3. 运行data_generator.py中的main()函数以生成人工数据集；
4. 运行metric_calculator.py中的main()函数以生成指标函数与可视化图表。