ScatterND
Description
Creates a copy of data and updates the data in the specified indices based on updates.
Input
Three inputs:
data: tensor of type float16 or float, of rank r>= 1.
indices: tensor of type int64, with rank q >= 1.
updates: tensor of type float16 or float, of rank = q + r - indices.shape[-1] - 1.
Output
One output:
y: tensor of the identical data type and shape as the input data.
Constraints
The index value of indices cannot be a negative number.
ONNX Opset Support
Opset v11/v12/v13/v14/v15/v16/v17/v18
Parent topic: Supported ONNX Operators