FilesLast commitLast update
7 days ago
7 months ago
20 days ago
10 months ago
7 days ago
3 months ago
1 month ago
11 days ago
README

Rfft1D

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: performs the RFFT (Fourier transform) computation on the input tensor. The output is a complex tensor that contains non-negative frequencies.

  • Computation formula:

    y=W⋅xy = W \cdot x

    Here, W is the DFT matrix, which is defined as Wjk=e−ij2πkNnW_{jk}=e^{-ij2\pi{\tfrac{k}{N}}n}

  • Example:

    Assume that x is {1, 2, 3, 4}. Then y = {10, 0, -2, 2, -2, 0} = {10 + 0j, -2 + 2j, -2 + 0j}

Parameter Description

Parameter Name Input/Output/Attribute Description Data Type Data Format
x Input Input tensor x in the formula. FLOAT ND
dft Input The input DFT matrix or twiddle matrix. With this matrix, the Rfft1D transform can be expressed as matrix multiplication. FLOAT ND
n Input Signal length. When the value range of n is [1, 4096], the DFT matrix should be passed in for dft. Otherwise, the twiddle matrix should be passed in. INT64 -
norm Input Normalization mode. Supported values: 1 indicates no normalization, 2 indicates normalization by 1/n, and 3 indicates normalization by 1/sqrt(n). INT64 -
y Output Output in the formula. FLOAT ND

Constraint Description

None

Invocation Description

Invocation Method Invocation Sample Description
aclnn invocation test_acl_rfft1_d Invokes the Rfft1D operator through the aclRfft1D API.