Engee documentation

Reshape

Changes the dimensionality of a vector or matrix signal.

reshape

Description

Block Reshape changes the dimension of the input signal to the dimension specified by the parameters Output dimensionality.

For example, you can convert a vector of length into a matrix of dimension 1 by or by 1.

Ports

Input

# IN_1 — input signal
scalar | vector | matrix | n-dimensional signal

Details

Input signal whose dimensions will be changed based on the Output dimensionality parameters.

Data types

Float16, Float32, Float64, Int8, Int16, Int32, Int64, Int128, UInt8, UInt16, UInt32, UInt64, UInt128, Bool, Fixed-point.

Complex numbers support

Yes

# Ref — output dimension
scalar | vector | matrix | n-dimensional signal

Details

Output signal dimensionality.

Dependencies

To use this port, set parameters Output dimensionality value Derive from reference input port.

Data types

Float16, Float32, Float64, Int8, Int16, Int32, Int64, Int128, UInt8, UInt16, UInt32, UInt64, UInt128, Bool, Fixed-point

Complex numbers support

Yes

Output

# OUT_1 — output signal
scalar | vector | matrix | n-dimensional signal

Details

Output signal with dimensions specified by the Output dimensionality parameters.

Data types

Float64, Int128, Fixed-point.

Complex numbers support

Yes

Parameters

Main

# Output dimensionality — output dimension
1-D array | Column vector (2-D) | Row vector (2-D) | Customize | Derive from reference input port

Details

The dimension of the output signal. Options for selection:

  • 1-D array - converts a vector, matrix, or multidimensional array to a vector (one-dimensional array). The output vector consists of the first column of the input matrix, followed by the second column, and so on (this option leaves the vector input signal unchanged).

  • Column vector (2-D) - converts a vector, matrix or multidimensional input signal into a column matrix of size by 1, where is the number of elements in the input signal. For matrices, the conversion is performed in column order. For multidimensional arrays the transformation is performed by the first dimension.

  • Row vector (2-D) - converts a vector, matrix or multidimensional input signal into a matrix-row of size 1 by , where is the number of elements in the input signal. For matrices, the conversion is performed in column order. For multidimensional arrays the transformation is performed by the first dimension.

  • Customize - converts the input signal into the output signal, the dimensions of which are set by the parameters Output dimensions.

  • Derive from reference input port - converts the input signal into an output signal whose dimensions are input to the Ref input port.

Values

1-D array | Column vector (2-D) | Row vector (2-D) | Customize | Derive from reference input port

Default value

1-D array

Program usage name

OutputDimensionality

Tunable

No

Evaluatable

No

# Output dimensions — output dimension
Tuple of Int64

Details

Dimensions for the output signal:

  • (N,) - for a vector output signal the parameter is specified as a tuple of one element defining the number of elements of the input vector.

  • (N, M) - for a matrix or multidimensional output signal, the parameter is specified as a tuple of n elements. The number of elements of the input signal must correspond to the number of elements specified by the value of the parameters in the input signal. In case of matrix signals, the data are selected from the columns of the input matrix and sequentially entered into the columns of the output matrix.

Dependencies

To use this parameter, set parameter Output dimensionality value Customize.

Default value

(1, 1)

Program usage name

OutputDimensions

Tunable

No

Evaluatable

Yes

Additional options

C code generation: Yes