Tutorials

Build a small argument map

from arglib.core import ArgumentGraph

graph = ArgumentGraph.new()
c1 = graph.add_claim("Exercise improves mood.")
c2 = graph.add_claim("People should exercise daily.")
graph.add_support(c1, c2, rationale="Mood improvement benefits health.")

Analyze structure

diagnostics = graph.diagnostics()
print(diagnostics["cycle_count"])
print(diagnostics["isolated_units"])

Save and reload

from arglib.io import save, load

save("graph.json", graph)
restored = load("graph.json", validate=True)

Export DOT for visualization

from arglib.viz import to_dot

dot = to_dot(graph)
print(dot)

Define an argument bundle (argument-as-subgraph)

bundle = graph.define_argument(
    [c1, c2],
    bundle_id="arg-1",
    metadata={"source": "doc-1"},
)
arg_graph = graph.to_argument_graph()

Attach evidence cards and supporting documents

from arglib.core import EvidenceCard, SupportingDocument

document = SupportingDocument(
    id="doc-1",
    name="Policy Report",
    type="pdf",
    url="https://example.com/policy.pdf",
)
graph.add_supporting_document(document)
card = EvidenceCard(
    id="ev-1",
    title="Evidence summary text.",
    supporting_doc_id=document.id,
    excerpt="Key finding...",
    confidence=0.6,
    metadata={"source_type": "report", "method": "expert"},
)
graph.add_evidence_card(card)
graph.attach_evidence_card(c1, card.id)

Propagate credibility

from arglib.reasoning import compute_credibility

cred = compute_credibility(graph)
print(cred.final_scores)

AI evaluation helpers

from arglib.ai import score_evidence, validate_edges

scores = score_evidence(graph)
edge_report = validate_edges(graph)

Long-document mining (split + reconcile)

from arglib.ai import LongDocumentMiner, SimpleArgumentMiner

miner = LongDocumentMiner(miner=SimpleArgumentMiner())
graph = miner.parse(long_text)