Charting nanocluster structures via convolutional neural networks
收藏资源简介：
The repository contains a notebook for the training of the autoencoder for the RDFs for structural classification. The notebook describes the procedure going from RDFs calculation to clustering of the reduced space. In the folder are contained Au147 structures, together with the associated pretrained AE, the 3D chart and the different clustering performed varying mean shift bandwidth. Files: - ChartAu147.ipynb: notebook - Configurations: directory with the dataset divided according to the CNA classification of the structures, xyz format with no headers, every 147 lines is a single structure - Libraries: directory with functions imported in the notebook - Precomputed: directory with the precomputed outputs - rdfs.npy: preocmputed RDFs of the data stored in configurations, npy format to load with NumPy - labels.npy: CNA labels of the RDFs, npy format to load with NumPy - model_au147.pth: pretrained model for au147 - scaler_au147.pkl: minmax scaler of the RDFs - chart_3d.dat: 3d space generated via the encoder on the au147 dataset - ae_reconstructions.npy: reconstructions of the rdfs of the model (model_au147.pth) - MSscanbw: pretrained mean shift clustering with different bandwidths, the file "clus_vs_bw.dat" reports the number of clusters associated to each bandwidth
本仓库包含用于结构分类的径向分布函数（Radial Distribution Function，RDF）自编码器（Autoencoder，AE）训练所用的Jupyter笔记本。该笔记本完整阐述了从径向分布函数计算到降维空间聚类的全流程。文件夹内包含Au147结构、配套的预训练自编码器模型、三维空间分布图，以及通过调整均值漂移（Mean Shift）带宽参数得到的多组聚类结果。 文件说明如下： - ChartAu147.ipynb：主运行笔记本 - Configurations：数据集目录，该目录下的数据集按照结构的公共邻居分析（Common Neighbor Analysis，CNA）分类规则进行划分，文件为无表头的xyz格式，每147行对应一个独立结构 - Libraries：存储笔记本中导入的自定义工具函数的目录 - Precomputed：预计算结果目录 - rdfs.npy：存储于Configurations目录下的数据集的预计算径向分布函数，为可通过NumPy加载的npy格式文件 - labels.npy：对应径向分布函数的CNA分类标签，为可通过NumPy加载的npy格式文件 - model_au147.pth：针对Au147体系的预训练自编码器模型 - scaler_au147.pkl：径向分布函数所用的最小最大缩放器 - chart_3d.dat：通过编码器对Au147数据集进行降维后生成的三维空间数据文件 - ae_reconstructions.npy：基于model_au147.pth模型对输入径向分布函数的重构结果文件 - MSscanbw：包含多组不同带宽参数的预训练均值漂移聚类模型，其中"clus_vs_bw.dat"文件记录了每组带宽参数对应的聚类簇数量
