Cumsum

Description

Computes the cumulative sum of the input tensor on the given axis. The cumulative sum is performed on the input elements along the given axis.

Input

• x: input tensor of type float16, float, or int32.
• axis: a 0-dimensional vector of type int32 or int64. The value range is [–r, r – 1]. The default value is 0. A negative value indicates that the dimension is calculated from the end. r indicates the number of dimensions of the input data x.

Attribute

• exclusive: The data type is int. The default value is 0, indicating that the sum of top-level elements is returned. If this parameter is set to 1, the sum of top-level elements is not returned.
• reverse: The data type is int. The default value is 0, indicating forward summation. If this parameter is set to 1, reverse summation is performed.

Output

y: output tensor of the identical data type as the input.

Constraints

None.

ONNX Opset Support

Opset v8/v9/v10/v11/v12/v13/v14/v15/v16/v17/v18