def test_count_alleles_subpops(self):
data = chunked.storage_registry['default'].array(diploid_genotype_data, chunklen=2)
g = GenotypeChunkedArray(data)
subpops = {'foo': [0, 2], 'bar': [1]}
ac_subpops = g.count_alleles_subpops(subpops)
for p in subpops.keys():
ac = g.take(subpops[p], axis=1).count_alleles()
aeq(ac, ac_subpops[p])
loc = np.array([True, False, True, False, True])
t = ac_subpops.compress(loc)
eq(3, len(t))
def filters_for_haplotyping(
genotypes: allel.GenotypeChunkedArray,
variants: allel.VariantChunkedTable,
chrom: str) -> (allel.GenotypeArray, allel.VariantTable):
""" Performs a series of filters to prepare the 'genotypes' and 'variants' object
for haplotyping.
Parameters:
genotypes (allel.GenotypeChunkedArray): GenotypesChunkedArray object.
variants (allel.VariantChunkedTable): VariantChunkedTable object.
chrom (str): What chromosome should be considered for the haplotype process.
Returns:
Tuple (allel.GenotypeArray, allel.VariantTable):
- allel.GenotypeArray: GenotypeArray object
- allel.VariantTable: VariantTable object
"""
# Filter by chrom
np_array_variants_in_chr = variants_filter_by_chrom(variants, chrom)
logger.debug(
"There are {count_variants_in_chr} variants in chromosome {chrom}".
format(
count_variants_in_chr=np.count_nonzero(np_array_variants_in_chr),
chrom=chrom))
# Filter by segregating SNPs
allele_count = genotypes.count_alleles()
np_array_log_sec = allele_count.is_segregating()
logger.debug("There are {count_log_sec} segregating SNPs".format(
count_log_sec=np.count_nonzero(np_array_log_sec)))
np_array_variants_to_keep = np_array_variants_in_chr & np_array_log_sec
logger.debug("Number of variants to keep {count_variants_to_keep}".format(
count_variants_to_keep=np.count_nonzero(np_array_variants_to_keep)))
# Subsets: perform the subset and load the results into memory uncompressed
genotypes_uc = genotypes.subset(np_array_variants_to_keep,
range(0, genotypes.n_samples))[:]
variants_np_array = variants[:]
variants_uc = variants_np_array.compress(np_array_variants_to_keep)
return genotypes_uc, variants_uc
def test_constructor(self):
# missing data arg
with assert_raises(TypeError):
# noinspection PyArgumentList
GenotypeChunkedArray()
# data has wrong dtype
data = 'foo bar'
with assert_raises(TypeError):
GenotypeChunkedArray(data)
# data has wrong dtype
data = np.array([4., 5., 3.7])
with assert_raises(TypeError):
GenotypeChunkedArray(data)
# data has wrong dimensions
data = np.array([1, 2, 3])
with assert_raises(TypeError):
GenotypeChunkedArray(data)
# data has wrong dimensions
data = np.array([[1, 2], [3, 4]]) # use HaplotypeChunkedArray instead
with assert_raises(TypeError):
GenotypeChunkedArray(data)
# diploid data (typed)
g = self.setup_instance(np.array(diploid_genotype_data, dtype='i1'))
aeq(diploid_genotype_data, g)
eq(np.int8, g.dtype)
# polyploid data (typed)
g = self.setup_instance(np.array(triploid_genotype_data, dtype='i1'))
aeq(triploid_genotype_data, g)
eq(np.int8, g.dtype)
def setup_instance(self, data, **kwargs):
data = chunked.storage_registry['default'].array(data, chunklen=2, **kwargs)
return GenotypeChunkedArray(data)
def setup_instance(self, data, dtype=None):
data = chunked.hdf5tmp_lzf_storage.array(data, dtype=dtype)
return GenotypeChunkedArray(data)
def setup_instance(self, data, **kwargs):
data = chunked.zarrtmp_storage.array(data, **kwargs)
return GenotypeChunkedArray(data)
