A graphical application for data normalization and mapping of heterogeneous data using the AMIS (Adaptive Multi-Interval Scale) method.
- 📊 Universal normalization for homogeneous and heterogeneous data
- ⚡ 5 AMIS models (3, 5, 9, 17 points + linear)
- 🎯 Adaptive model selection (NEW in v4.3) - automatically adjusts to data volume
- 🔄 Cross-metric analysis - compare different measurement units
- 📈 Interactive visualization and CSV export
- 🎯 Switch between original/normalized data views
- ✅ Input validation with clear error messages
AMIS v4.3 automatically selects available normalization models:
- 10-19 points: Linear + 3-point model
- 20-49 points: + 5-point model
- 50-99 points: + 9-point model
- 100+ points: All 5 models available
The interface adapts automatically - unavailable models are disabled in dialogs.
Prerequisites: Python 3.8+ (pandas, numpy, scipy, matplotlib)
git clone https://github.com/Famimot/AMIS_Normalization_Tool
cd AMIS_Normalization_Tool
pip install -r requirements.txt
pip install -e .
From source (the package is currently under review for publication):
# 1. Download the source code and navigate to the AMIS folder
# 2. Install the package and its dependencies
pip install -e .Note: The pip install amis_tool command will become available after the package is published on PyPI.
After installation, launch the graphical interface with the following command:
python -m amis_toolExpected behavior: A splash screen will appear briefly, followed by the main application window.
The program requires data in two-column format:
Country,GDP
Japan,4026210821147
Germany,4659929336891
China,18743803170827
EU,19423000000000
United States,29184890000000
Requirements:
- First column: Labels/identifiers (country names, student IDs, item names)
- Second column: Numerical values to normalize (grades, GDP, scores, etc.)
- Minimum: 10 valid numerical values
- Formats supported: Excel (.xlsx, .xls) and CSV (.csv)
Note: The first column labels are preserved during normalization and appear in output tables.
| File | Format | Type | Size | Available Models* |
|---|---|---|---|---|
World_Bank_Nominal_GDP_All_Countries_2024.xlsx |
Excel | Country GDP | 189 rows | All 5 models |
World_Bank_Nominal_GDP_87_Countries_2024.xlsx |
Excel | Country GDP | 87 rows | Linear, 3, 5, 9-point |
Student_Grades_History_Grade11_Raw_Data.xlsx |
Excel | Educational grades | 879 rows | All 5 models |
Student_Grades_History_Grade11_Raw_Data.csv |
CSV | Educational grades | 879 rows | All 5 models |
*Available models automatically adjust based on row count (10+ points required)
Usage: python -m amis_tool → File → Open → Select file → Choose data type according to table above.
Figure 1: AMIS main interface showing simultaneous normalization of heterogeneous datasets: Russian school history grades (left panel) and World Bank GDP 2024 (right panel).
Figure 2: Four-panel comparative visualization:
- Top-left: AMIS normalization curve for history grades
- Top-right: AMIS normalization curve for country GDP
- Bottom-left: Correspondence mapping between grade and GDP AMIS scales
- Bottom-right: Direct comparison of both metrics on unified 0-100 AMIS scale
AMIS/
├── amis_tool/ # Main Python package
│ ├── __init__.py
│ ├── __main__.py # Main entry point
│ ├── core/
│ │ └── amis_calculations.py
│ ├── gui/
│ │ ├── main_window.py
│ │ ├── dialogs.py
│ │ └── widgets.py
│ └── utils/
│ └── helpers.py
├── examples/ # Example datasets
│ ├── Student_Grades_History_Grade11_Raw_Data.xlsx
│ ├── Student_Grades_History_Grade11_Raw_Data.csv
│ ├── World_Bank_Nominal_GDP_All_Countries_2024.xlsx
│ └── World_Bank_Nominal_GDP_87_Countries_2024.xlsx
├── images/ # Application screenshots for documentation
│ ├── screenshot-main.png
│ └── screenshot-normalization-and-mapping.png
├── paper.md # JOSS article
├── paper.bib # Bibliography (with preprint)
├── setup.py # Installation configuration
├── requirements.txt # Python dependencies
├── run_amis.py # Alternative launcher script
├── README.md # Documentation
├── LICENSE # MIT License
├── .gitignore # Git ignore rules
└── .gitattributes # Git attributes
If you use AMIS in your research or software in your projects, please cite both the methodology papers and the respective replication datasets:
- Main Theory & Application: Kravtsov, G. G. (2026). Adaptive Multi-Interval Scale (AMIS): Construction of a Unified Metric Space for Heterogeneous Data. Modelling and Data Analysis, 16(2), 67–83. https://doi.org/10.17759/mda.2026160203
- Algorithm & IT Framework: Kravtsov, G. G. (2026). Adaptive Multi-Interval Scale (AMIS): Algorithm for normalization of aggregated data in a unified measurement space. Research Result. Information Technologies, 11(2), 50–61. https://doi.org/10.18413/2518-1092-2026-11-2-0-5
If you use the benchmark datasets for verification or reproduction, please cite them as follows:
- Graphical Visualization Data: Kravtsov, G. G. (2026). Replication Data for: Adaptive Multi-Interval Scale (AMIS): Graphical Visualization of a Unified Measurement Space [Data set]. Harvard Dataverse. https://doi.org/10.7910/DVN/HXQYRR
- Medical Device Data: Kravtsov, G. G. (2026). Supplementary Materials for the Article "Adaptive Multi-Interval Scale (AMIS): An Algorithm for Normalizing Medical Device Data into a Unified Metric Space" [Data set]. Harvard Dataverse. https://doi.org/10.7910/DVN/OTEUVU
- SSRN: Kravtsov, G. G. (2026). Adaptive Multi-interval Scale (AMIS): Graphical Visualization of a Unified Measurement Space. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.7090960
- OSF: Kravtsov, G. G. (2025). Universal Adaptive Normalization Scale (AMIS): A methodology for integrating heterogeneous social and educational metrics. OSF Preprints. https://doi.org/10.17605/OSF.IO/BDT2K
[@kravtsov2025amistool]