def test_parquet_family_bin(fam1, fam2, gt):
sv = SummaryVariant(summary_alleles_chr1)
fv1 = FamilyVariant(sv, fam1, gt, None)
fv2 = FamilyVariant(sv, fam2, gt, None)
family_bin_size = 10
pd = ParquetPartitionDescriptor(["1"], 1000, family_bin_size)
for fa1, fa2 in zip(fv1.alleles, fv2.alleles):
assert pd._evaluate_family_bin(fa1) == 9
assert pd._evaluate_family_bin(fa2) == 6
assert (pd.variant_filename(fa1) == "region_bin=1_11/family_bin=9/"
"variants_region_bin_1_11_family_bin_9.parquet")
assert (pd.variant_filename(fa2) == "region_bin=1_11/family_bin=6/"
"variants_region_bin_1_11_family_bin_6.parquet")
def test_parquet_region_bin(fam1, gt, chromosomes, region_length,
summary_alleles, expected):
sv = SummaryVariant(summary_alleles)
fv = FamilyVariant(sv, fam1, gt, None)
pd = ParquetPartitionDescriptor(chromosomes, region_length)
region_bin = pd._evaluate_region_bin(fv)
for fa in fv.alleles:
assert region_bin == expected
assert (pd.variant_filename(fa) ==
f"region_bin={region_bin}/variants_region_bin_{region_bin}"
f".parquet")
def test_parquet_frequency_bin(fam1, gt, attributes, rare_boundary, expected):
summary_alleles = [
SummaryAllele("1", 11539, "T", None, 0, 0, attributes=attributes)
] * 3
sv = SummaryVariant(summary_alleles)
fv = FamilyVariant(sv, fam1, gt, None)
pd = ParquetPartitionDescriptor(["1"], 1000, rare_boundary=rare_boundary)
for fa in fv.alleles:
assert pd._evaluate_frequency_bin(fa) == expected
assert (pd.variant_filename(
fa) == f"region_bin=1_11/frequency_bin={expected}/" +
f"variants_region_bin_1_11_frequency_bin_{expected}.parquet")
def test_parquet_coding_bin(fam1, gt, eff1, eff2, eff3, coding_effect_types,
expected):
summary_alleles = [
SummaryAllele("1", 11539, "T", None, 0, 0),
SummaryAllele("1", 11539, "T", "G", 0, 1, attributes={"effects":
eff1}),
SummaryAllele("1", 11539, "T", "C", 0, 2, attributes={"effects":
eff2}),
SummaryAllele("1", 11539, "T", "A", 0, 3, attributes={"effects":
eff3}),
]
gt = np.array([[0, 1, 0], [2, 0, 3]], dtype="int8")
sv = SummaryVariant(summary_alleles)
fv = FamilyVariant(sv, fam1, gt, None)
pd = ParquetPartitionDescriptor(["1"],
1000,
coding_effect_types=coding_effect_types)
for fa, ex in zip(fv.alleles, expected):
assert pd._evaluate_coding_bin(fa) == ex
assert (
pd.variant_filename(fa) == f"region_bin=1_11/coding_bin={ex}/" +
f"variants_region_bin_1_11_coding_bin_{ex}.parquet")