"""Contract tests for core AnnData utility helpers."""
from __future__ import annotations
import numpy as np
import pytest
from scipy import sparse
from chatspatial.utils.adata_utils import (
get_raw_data_source,
get_spatial_key,
require_spatial_coords,
to_dense,
)
from chatspatial.utils.exceptions import DataError
def test_get_spatial_key_returns_known_key(minimal_spatial_adata):
adata = minimal_spatial_adata.copy()
assert get_spatial_key(adata) in adata.obsm
def test_require_spatial_coords_falls_back_to_obs_xy(minimal_spatial_adata):
adata = minimal_spatial_adata.copy()
del adata.obsm["spatial"]
adata.obs["x"] = np.arange(adata.n_obs, dtype=float)
adata.obs["y"] = np.arange(adata.n_obs, dtype=float) + 1.0
coords = require_spatial_coords(adata)
assert coords.shape == (adata.n_obs, 2)
assert np.allclose(coords[:, 1], coords[:, 0] + 1.0)
def test_require_spatial_coords_rejects_identical_points(minimal_spatial_adata):
adata = minimal_spatial_adata.copy()
adata.obsm["spatial"] = np.ones((adata.n_obs, 2), dtype=float)
with pytest.raises(DataError, match="identical"):
require_spatial_coords(adata)
def test_get_raw_data_source_prefers_raw_for_complete_genes(minimal_spatial_adata):
adata = minimal_spatial_adata.copy()
adata.raw = adata.copy()
adata = adata[:, :5].copy()
out = get_raw_data_source(adata, prefer_complete_genes=True)
assert out.source == "raw"
assert len(out.var_names) == 24
def test_get_raw_data_source_uses_counts_when_not_preferring_raw(minimal_spatial_adata):
adata = minimal_spatial_adata.copy()
adata.raw = adata.copy()
adata.layers["counts"] = np.asarray(adata.X).astype(np.int64)
out = get_raw_data_source(adata, prefer_complete_genes=False)
assert out.source == "counts_layer"
assert len(out.var_names) == adata.n_vars
def test_get_raw_data_source_requires_integer_counts(minimal_spatial_adata):
adata = minimal_spatial_adata.copy()
adata.X = (np.asarray(adata.X, dtype=float) + 0.25).astype(np.float64)
if "counts" in adata.layers:
del adata.layers["counts"]
with pytest.raises(DataError, match="No raw integer counts found"):
get_raw_data_source(
adata,
prefer_complete_genes=False,
require_integer_counts=True,
)
def test_to_dense_handles_sparse_and_dense_copy_semantics():
dense = np.arange(6, dtype=float).reshape(2, 3)
sparse_x = sparse.csr_matrix(dense)
out_sparse = to_dense(sparse_x)
assert isinstance(out_sparse, np.ndarray)
assert np.array_equal(out_sparse, dense)
out_copy = to_dense(dense, copy=True)
out_copy[0, 0] = -1
assert dense[0, 0] == 0
