Adaptive Multi-Interval Scale (AMIS): Open-Source Software, Source Data and Results for Normalizing and Comparing Raw & Aggregated Metrics

This dataset contains the complete resources for applying the Adaptive Multi-Interval Scale (AMIS) method to normalize and compare heterogeneous data. The AMIS method addresses the fundamental problem of integrating indicators measured in fundamentally incomparable scales. It is specifically designed to work with both raw individual measurements and aggregated data (frequency distributions, summary tables). The dataset provides open-source Python software that implements the full AMIS pipeline: from data loading and transformation to adaptive normalization and the creation of matching tables. It includes tested algorithms for converting aggregated data into a normalized form and a robust inverse transformation mechanism. The dataset ensures full reproducibility of the research and provides a practical toolkit for interdisciplinary data integration and analysis.

本数据集包含应用自适应多区间量表（Adaptive Multi-Interval Scale, AMIS）方法对异构数据进行归一化与比较所需的全部资源。该AMIS方法可解决核心难题：对度量尺度本质上不可通约的各类指标进行整合，其专门适配原始个体测量数据与聚合类数据（包括频数分布、汇总表格）两类数据形式。本数据集附带可实现完整AMIS全流程的开源Python软件，涵盖数据加载、数据转换、自适应归一化直至匹配表生成的全部环节，其中包含经验证的聚合数据归一化转换算法与鲁棒的逆变换机制。本数据集可保障研究的完全可复现性，并为跨学科数据整合与分析提供实用工具集。