def compute(self, today, assets, out, value):
if isinstance(value, LabelArray):
out[:] = ~np_any(value.is_missing(), axis=0)
else:
out[:] = ~np_any(
is_missing(value, self.inputs[0].missing_value),
axis=0,
)
def compute(self, today, assets, out, value):
if isinstance(value, LabelArray):
out[:] = ~np_any(value.is_missing(), axis=0)
else:
out[:] = ~np_any(
is_missing(value, self.inputs[0].missing_value),
axis=0,
)
def _compute(self, arrays, dates, assets, mask):
data = arrays[0]
group_labels, null_label = self.inputs[1]._to_integral(arrays[1])
effective_mask = (
mask
& (group_labels != null_label)
& ~is_missing(data, self.inputs[0].missing_value)).view(uint8)
return grouped_masked_is_maximal(
# Unconditionally view the data as int64.
# This is safe because casting from float64 to int64 is an
# order-preserving operation.
data.view(int64_dtype),
# PERF: Consider supporting different sizes of group labels.
group_labels.astype(int64_dtype),
effective_mask,
)
def _compute(self, arrays, dates, assets, mask):
data = arrays[0]
group_labels, null_label = self.inputs[1]._to_integral(arrays[1])
effective_mask = (
mask
& (group_labels != null_label)
& ~is_missing(data, self.inputs[0].missing_value)
).view(uint8)
return grouped_masked_is_maximal(
# Unconditionally view the data as int64.
# This is safe because casting from float64 to int64 is an
# order-preserving operation.
data.view(int64_dtype),
# PERF: Consider supporting different sizes of group labels.
group_labels.astype(int64_dtype),
effective_mask,
)
def _compute(self, arrays, dates, assets, mask):
data = arrays[0]
if isinstance(data, LabelArray):
return ~data.is_missing()
return ~is_missing(arrays[0], self.inputs[0].missing_value)
def _compute(self, arrays, dates, assets, mask):
data = arrays[0]
if isinstance(data, LabelArray):
return data.is_missing()
return is_missing(arrays[0], self.inputs[0].missing_value)
