polli-typus 0.4.3

Taxonomic types, projections and async taxonomy services for Polli-Labs.

Typus

Shared taxonomy & geo‑temporal types for the Polli‑Labs ecological stack

Documentation: https://docs.polli.ai/typus

Typus centralises every domain object that the rest of our platform — linnaeus, pollinalysis-server, dashboards … — needs: taxon records, clades, hierarchical classification results, projection helpers and async database services. Anything that speaks taxonomy imports Typus and stays DRY.

---

Features

• Wide ancestry view – expanded_taxa ORM exposes each rank (L10 → L70) on a single row for constant‑time lineage queries.
• Async services – PostgresTaxonomyService (ltree) & SQLiteTaxonomyService (fixture) share one interface.
• Pydantic v2 models – Taxon, Clade, HierarchicalClassificationResult, all JSON‑Schema‑exportable.
• Taxonomy summaries & pollinator groups – TaxonSummary trails plus coarse PollinatorGroup helpers for UI labels.
• Projection utils – lat/lon ↔ unit‑sphere, cyclical‑time features, multi‑scale elevation sinusoids.
• Optional drivers only when you need them – install polli-typus[postgres] or [sqlite]; core install stays lightweight.
• Offline SQLite loader – typus-load-sqlite CLI builds and caches the offline dataset

---

Requirements

• Python ≥ 3.10

---

Installation

Core (no DB drivers)

uv pip install polli-typus        # import typus

With Postgres backend

uv pip install "polli-typus[postgres]"    # adds asyncpg

With SQLite only (CI, offline, sandboxes)

uv pip install "polli-typus[sqlite]"

Development / tests / lint

uv pip install -e ".[dev,sqlite]"   # pytest, pytest-asyncio, ruff, pre-commit, aiosqlite …

Install with plain pip:

pip install -e ".[dev,sqlite]"

---

Quick start

SQLite (recommended for getting started)

# Download or build the expanded taxonomy dataset locally (~475MB sqlite / ~440MB tsv.gz)
# The loader creates recommended indexes by default for fast name search
typus-load-sqlite --sqlite expanded_taxa.sqlite

from pathlib import Path
from typus.services import SQLiteTaxonomyService

svc = SQLiteTaxonomyService(Path("expanded_taxa.sqlite"))
bee = await svc.get_taxon(630955)           # Anthophila
print(bee.scientific_name, bee.rank_level)  # Anthophila RankLevel.L32

You can also load on-demand in code (will download if missing):

from pathlib import Path
from typus.services import load_expanded_taxa, SQLiteTaxonomyService

db_path = load_expanded_taxa(Path("expanded_taxa.sqlite"))  # create_indexes=True by default
svc = SQLiteTaxonomyService(db_path)

Name search (v0.4.0+)

# Scientific prefix match
taxa = await svc.search_taxa("Apis", scopes={"scientific"}, match="prefix")

# Vernacular (common name) exact
taxa = await svc.search_taxa("honey bee", scopes={"vernacular"}, match="exact")

Taxonomy summaries & pollinator groups (v0.4.2)

from pathlib import Path
from typus import PollinatorGroup
from typus.services import SQLiteTaxonomyService

svc = SQLiteTaxonomyService(Path("expanded_taxa.sqlite"))

# Compact trail for UIs
bee_summary = await svc.taxon_summary(630955)  # Anthophila
print(bee_summary.format_trail())  # Animalia → Arthropoda → Insecta → Hymenoptera → Apidae → Anthophila

# Coarse pollinator grouping
groups = await svc.pollinator_groups_for_taxon(47219)  # Apis mellifera
assert PollinatorGroup.BEE in groups

Elevation (Postgres only)

import os
from typus import PostgresRasterElevation

dsn = os.getenv("ELEVATION_DSN") or os.getenv("TYPUS_TEST_DSN")
elev = PostgresRasterElevation(dsn, raster_table=os.getenv("ELEVATION_TABLE", "elevation_raster"))

la = await elev.elevation(34.0522, -118.2437)
vals = await elev.elevations([
    (34.0522, -118.2437),  # LA
    (0.0, -30.0),          # ocean (likely None)
])

Geo helpers

from typus import latlon_to_unit_sphere
print(latlon_to_unit_sphere(31.5, -110.4))  # → x, y, z on S²

Postgres (optional for production deployments)

from typus import PostgresTaxonomyService
svc = PostgresTaxonomyService("postgresql+asyncpg://user:pw@host/db")
bee = await svc.get_taxon(630955)

---

Developer guide

• Lint & tests (one‑liner)

  ruff format . && ruff check . && pytest -q
  

• Format whole repo

  ruff format .
  

• JSON Schemas – python -m typus.export_schemas → typus/schemas/

• SQLite fixture – python scripts/gen_fixture_sqlite.py

• Pre‑commit hooks – pre-commit install

Environment Variables

• TYPUS_TEST_DSN: Postgres DSN for tests and perf harness (e.g., postgresql+asyncpg://user:pw@host/db).
• POSTGRES_DSN: Alternate Postgres DSN; used if TYPUS_TEST_DSN is unset.
• ELEVATION_DSN: Optional DSN override for elevation tests; falls back to TYPUS_TEST_DSN.
• ELEVATION_TABLE: Elevation raster table name (default: elevation_raster).
• TYPUS_ELEVATION_TEST: Set 1 to enable guarded elevation tests.
• Perf harness:
  • TYPUS_PERF_WRITE=1: write report to dev/agents/perf_report.md.
  • TYPUS_PERF_VERIFY=1: enable result sanity checks.
  • TYPUS_PERF_EXPLAIN=1: append PG EXPLAIN snippets.

---

Publishing (maintainers)

See build/typus_publish.md for tag → TestPyPI → PyPI workflow.

---

License

MIT © 2025 Polli Labs