Tools: computational analysis
Networks, machine learning, embeddings, and simulation
The maintained version of Chapter 8’s tool list: networks, machine learning, embeddings, and simulation.
Use these tools to implement the network, prediction, and simulation exercises in Chapter 8. Choose a method from your research question, then check whether the tool supports the assumptions and diagnostics you need. Tool selection reviewed August 2026.
Network analysis
- NetworkX (Python): graph construction, measures, and algorithms; used in the worked network notebook.
- igraph (C, Python, R): graph analysis with a compiled core and interfaces in several languages.
- statnet (R): ERGMs and statistical models of tie formation, for explanatory work on network structure.
- graph-tool (Python): graph analysis with a compiled core; check its installation requirements.
Visualization: Gephi for interactive network layouts, Cytoscape for attribute-rich networks. Always report the layout algorithm and its parameters (Chapter 9, §9.4.5).
Machine learning
- scikit-learn: classifiers, preprocessing, model selection, and evaluation metrics.
- XGBoost and LightGBM: gradient boosting for tasks such as prediction from tabular data. Compare performance with a simpler baseline on held-out observations.
- SHAP: feature attributions, global and per-prediction. Remember Chapter 7’s caution: an explanation of a prediction is not a causal claim.
Embeddings
- gensim: Word2Vec and related vector models.
- Sentence-Transformers: embeddings for sentences and other text units.
- UMAP: projection for visualization. Report the algorithm, its parameters, and the output dimensionality; Chapter 9 (§9.4.5) shows how much the picture changes without them.
Simulation
Causal inference with machine learning
Reproducibility
Set and record random seeds, and compare results across seeds when variation matters to your conclusion. Record dependencies with a lock file (Chapter 9, §9.3.2). See communication tools for environment management.