FilesLast commitLast update
13 days ago
3 months ago
23 days ago
2 months ago
7 days ago
2 months ago
7 days ago
3 months ago
1 month ago
14 days ago
README

Cholesky

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: computes the Cholesky decomposition of a real symmetric positive-definite matrix.

  • Computation formula:

    self=out×outHself = out \times out^{H}

Parameter Description

Parameter Name Input/Output/Attribute Description Data Type Data Format
self Input Input tensor, which is self in the formula. FLOAT ND
upper Attribute Specifies whether to perform the upper-triangular or lower-triangular matrix decomposition. BOOL -
out Output Output tensor, which is out in the formula. FLOAT ND

Constraint Description

  • The dimensions of self must be greater than or equal to 2, and self must be a positive-definite matrix.

Invocation Description

Invocation Method Invocation Sample Description
aclnn invocation test_aclnn_linalg_cholesky Invokes the Cholesky operator through the aclnnLinalgCholesky API.