FloorMod

English | 简体中文

Product Support Status

Product	Supported
Ascend 950PR&950DT products	√
Atlas A3 products	√
Atlas A2 products	√
Atlas 200I/500 A2 inference products	×
Atlas inference products	×
Atlas training products	√

Function Description

• Operator function: broadcasts the scalar self into a tensor with the same shape as the tensor other, and then converts each element into the remainder obtained by dividing it by the corresponding element of other. The result has the same sign as the divisor other, and the absolute value of the result is less than the absolute value of other. The actual computation of remainder(self, other) is equivalent to the following formula:

  $$ou{t}_i=self-floor(self/othe{r}_i)\ast othe{r}_i$$

• Example:

self = 5.0   # float
other = tensor([[-1, -2],
                [-3, -4]]).type(int32)
result = remainder(self, other)

# Value of result
# tensor([[ 0., -1.],
#         [-1., -3.]])  float

# For the element -4 in other, the computation result is 5 - floor(5 / -4) * -4 = -3
# It can be seen that the absolute value of the final result -3 is less than the absolute value of the original -4.

Parameter Description

Parameter Name	Input/Output/Attribute	Description	Data Type	Data Format
x1	Input	Input tensor self_i in the formula	FLOAT16, BFLOAT16, FLOAT, INT32, INT64	ND
x2	Input	Input tensor other_i in the formula	FLOAT16, BFLOAT16, FLOAT, INT32, INT64	ND
y	Output	Output tensor out_i in the formula	FLOAT16, BFLOAT16, FLOAT, INT32, INT64	ND

Invocation Description

Invocation Method	Invocation Sample	Description
aclnn invocation	test_aclnn_remainder_scalar_tensor	Invokes the FloorMod operator through the aclnnRemainderScalarTensor API