Examples

Runnable demos live in examples/. Activate the environment first (. .venv/bin/activate).

Python demos

Demo	What it shows
family_tre.py	Pattern-directed forward chaining with TRE — deriving family relationships from rules
belief_revision_ltre.py	Assumptions and automatic belief revision in LTRE (now a thin .kb loader)
coloring_dds.py	Graph coloring solved by dependency-directed search with nogood learning

python examples/family_tre.py
python examples/coloring_dds.py

World models as .kb files

The .kb DSL lets a world model live in a data file, separate from Python. Each file is self-checking: its expect lines assert the resulting belief state, and pytest tests/test_kb_files.py runs every one of them.

File	Demonstrates
belief_revision.kb	Retractable assumptions and belief revision
modus_tollens.kb	Contrapositive reasoning via BCP
taxonomy.kb	The taxonomy (exactly-one) directive
completeness.kb	The complete directive (prime-implicate completion)
diagnosis.kb	A small diagnosis model
family.kb	Implication rules over family relationships

Run any .kb file directly:

python examples/run_kb.py examples/kb/belief_revision.kb

A minimal model looks like this:

# examples/kb/belief_revision.kb
rain         -> wet ground
sprinkler on -> wet ground
assume rain
expect wet ground true

The full directive set (assert, assume/retract, -> rules, taxonomy, complete, contradiction, query, expect) is documented in the ltms.dsl source. Chapter-specific .kb models also live under exercises/chNN/kb/ — see the Exercises page.