def _get_sample_gts(gts, sample_i, sample_j, indi_cache):
if sample_i in indi_cache:
indi1, is_missing_1 = indi_cache[sample_i]
else:
indi1 = gts[:, sample_i]
is_missing_1 = is_missing(indi1)
indi_cache[sample_i] = indi1, is_missing_1
if sample_j in indi_cache:
indi2, is_missing_2 = indi_cache[sample_j]
else:
indi2 = gts[:, sample_j]
is_missing_2 = is_missing(indi2)
indi_cache[sample_j] = indi2, is_missing_2
is_called = numpy.logical_not(numpy.logical_or(is_missing_1, is_missing_2))
indi1 = indi1[is_called]
indi2 = indi2[is_called]
assert issubclass(indi1.dtype.type, numpy.integer)
assert issubclass(indi2.dtype.type, numpy.integer)
return indi1, indi2
def _get_sample_gts(gts, sample_i, sample_j, indi_cache):
if sample_i in indi_cache:
indi1, is_missing_1 = indi_cache[sample_i]
else:
indi1 = gts[:, sample_i]
is_missing_1 = is_missing(indi1)
indi_cache[sample_i] = indi1, is_missing_1
if sample_j in indi_cache:
indi2, is_missing_2 = indi_cache[sample_j]
else:
indi2 = gts[:, sample_j]
is_missing_2 = is_missing(indi2)
indi_cache[sample_j] = indi2, is_missing_2
is_called = numpy.logical_not(numpy.logical_or(is_missing_1, is_missing_2))
indi1 = indi1[is_called]
indi2 = indi2[is_called]
assert issubclass(indi1.dtype.type, numpy.integer)
assert issubclass(indi2.dtype.type, numpy.integer)
return indi1, indi2
def _snp_has_enough_data(variations, max_missing_rate_in_ref_snp=0.1):
gts = variations[GT_FIELD]
missing_rate = numpy.sum(is_missing(gts, axis=2), axis=1) / gts.shape[1]
return missing_rate < max_missing_rate_in_ref_snp
