How otniel113/segmentasipelanggan works
Project type and purpose
• This is a data-analysis/research project delivered as a single Jupyter Notebook (Segmentasi_Pelanggan.ipynb).
• Its purpose is to perform customer segmentation (clustering) as part of a book-chapter publication.
File structure
• Top-level items:
• README.md – high-level description.
• Segmentasi_Pelanggan.ipynb – the core code and exposition.
• No separate “src” or “data” folders; everything lives in one notebook.
Extract insights from the README
• Confirms it’s research code accompanying a book chapter.
• No extra architectural notes—so we infer a linear data pipeline inside the notebook.
a. Main components
• Data sources (e.g., CSV files, database tables, or mock data loaded into the notebook)
• Notebook environment (Jupyter server + Python kernel)
• ETL/Preprocessing module (Pandas transformations)
• Clustering algorithm component (e.g., scikit-learn’s KMeans)
• Visualization & reporting (Matplotlib, Seaborn, output cells)
• External dependencies (Python libraries, conda/pip environment)
b. Relationships and interactions
1. Data sources → Notebook (data ingestion)
2. Notebook → ETL component (preprocessing steps)
3. ETL → Clustering component (model fitting)
4. Clustering → Visualization (plots, tables)
5. Notebook → Outputs (notebook file, exported charts, tables)
c. Architectural patterns/design principles
• Linear data-pipeline pattern: a sequence of steps from raw data to insights.
• Single “process host” (the notebook) orchestrates all stages.
d. Key technologies/libraries
• Jupyter Notebook interface
• Python environment (virtualenv/conda)
• pandas (data manipulation)
• scikit-learn (clustering)
• matplotlib / seaborn (visualization)
Detail without overcomplexity
• Every logical step in the notebook as its own component.
• Dependencies on external libraries and data sources.
• Don’t invent extra services—stick strictly to what’s in the notebook and its environment.