define a placeholder for a Tensor
create a Tensor filled with ones
create a Tensor filled with zeros
list, dimensions of the output Tensor
(optional) options as specified below
datatype=one of string,truefalse,integer,integer,float,float,complex, or complex
The value of option datatype specifies the type of data this Tensor will hold. If a Tensor t was provided as input, the default value of datatype is the datatype of t; otherwise, the default is float.
name = string
The value of option name specifies an optional name for this Tensor, to be displayed in output and when visualizing the dataflow graph.
The Fill(shape,val,opts) command creates a Tensor in the active dataflow graph with shape shape filled with the scalar value val.
The Ones(shape,opts) and Zeros(shape,opts) commands behave similarly but create Tensors whose whose entries are 1 and 0, respectively. If x is a Tensor, the val of the resulting Tensor matches x.
If t is a Tensor, the commands Ones(t,opts) and Zeros(t,opts) create Tensors whose whose shape is identical to t but whose entries are 1 and 0, respectively.
These functions are part of the DeepLearning package, so they can can be used in the short form Fill(..) only after executing the command with(DeepLearning). However, can always be accessed through the long form of the command by using DeepLearning[Fill](..).
T1 ≔ Fill⁡2,3,0.75
T1≔DeepLearning TensorShape: [2, 3]Data Type: float
T2 ≔ Ones⁡T1
T2≔DeepLearning TensorShape: [2, 3]Data Type: float
T3 ≔ Zeros⁡3,4,datatype=float8
T3≔DeepLearning TensorShape: [3, 4]Data Type: float
The DeepLearning[Fill], DeepLearning[Ones] and DeepLearning[Zeros] commands were introduced in Maple 2018.
For more information on Maple 2018 changes, see Updates in Maple 2018.
Download Help Document