Real

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: returns the real part of each element in the input tensor.

• Computation formula:

  $$outpu{t}_i=\mathrm{R}\mathrm{e}(inpu{t}_i)$$

  where:

  • When $inpu{t}_i=a_i+b_i\cdot j$ is a complex number, $outpu{t}_i=a_i$;
  • When $inpu{t}_i$ is a real number, $outpu{t}_i=inpu{t}_i$ (pass-through).

Parameter Description

Parameter Name	Input/Output/Attribute	Description	Data Type	Data Format
input	Input	Input tensor for real part extraction, which is $input_i$ in the formula.	FLOAT, FLOAT16, COMPLEX64, COMPLEX32	ND
Tout	Optional Attribute	Expected output dtype enumeration value. The default value is DT_FLOAT. The actual output type is deduced from the input dtype.	INT	-
output	Output	Output tensor after real part extraction, which is $output_i$ in the formula. The shape is the same as that of input.	FLOAT, FLOAT16	ND

Correspondence between input and output dtypes:

input	output
FLOAT	FLOAT
FLOAT16	FLOAT16
COMPLEX64	FLOAT
COMPLEX32	FLOAT16

Constraint Description

None

Invocation Description

Invocation Method	Invocation Sample	Description
aclnn invocation	test_aclnn_real	Invokes the Real operator through the aclnnReal API.
Graph-mode invocation	test_geir_real	Invokes the Real operator by building a graph through the operator IR.