Sparse Matrix-Vector Product

Multiply sparse matrix with input vector

Since R2024b

• 

Libraries:
HDL Coder / RCP and HIL

Description

The Sparse Matrix-Vector Product block multiplies the sparse matrix A with an input vector x.

Examples

Model 2-D Matrix-Vector Multiplication

This example builds a 2-dimensional matrix-vector multiplication function by using the Sparse Matrix-Vector Product block.

To implement this function, specify the parameter Matrix A as [1,2,0;3,0,4;0,5,6] and input vector x as [1,2,3] for the model.

load_system("SMVPblock2DExample")
sim("SMVPblock2DExample");
open_system("SMVPblock2DExample");

You can verify the results by performing 2-dimensional matrix-vector multiplication. In the MATLAB® Command Window, enter the value of matrix A and input vector x:

A = [1,2,0;3,0,4;0,5,6];
x = [1,2,3];

Compute x' and multiply with matrix A:

x = x';
y = A*x

y =

     5
    15
    28

Model 3-D Matrix-Vector Multiplication

This example builds a 3-dimensional matrix-vector multiplication function by using the Sparse Matrix-Vector Product block.

To implement this function, specify the parameter Matrix A as:

A(:,:,1) = [1,2,0;3,0,4;0,5,6]; A(:,:,2) = [1,2,3;3,5,4;7,5,6]; A(:,:,3) = [1,2,0;3,0,4;0,5,6];

and the input vector x as [1,2,3]. There is a Selector block to select the required matrix configuration based on specified indices. The Selector block has zero-based index mode where an index of 0 specifies the first element of the input vector. Therefore, for a 3-D input signal, 0, 1, and 2 are the valid input elements. In this example, the value of input for Sel is 2. For more information, see Selector.

load_system("SMVPblock3DExample")
sim("SMVPblock3DExample");
open_system("SMVPblock3DExample");

To perform 3-dimensional matrix-vector multiplication, in the MATLAB® Command Window, enter the value of matrix A and input vector x:

A(:,:,1) = [1,2,0;3,0,4;0,5,6];
A(:,:,2) = [1,2,3;3,5,4;7,5,6];
A(:,:,3) = [1,2,0;3,0,4;0,5,6];
x = [1,2,3];

For the 3-D input signal, the value of input signal Sel = 2 selects matrix A(:,:,3). Compute x' and multiply with matrix A(:,:,3):

x = x';
y = A(:,:,3)*x

y =

     5
    15
    28

Ports

Input

x — Input signal

scalar | vector

Input signal, specified as a scalar or vector. The Sparse Matrix-Vector Product block supports fixed-point data types. The number of input ports depends on the dimension of Matrix A. For example, see Model 3-D Matrix-Vector Multiplication.

Data Types: single | double | fixed point

Output

y — Output signal

scalar | vector

Output signal, returned as a scalar or vector. The output signal is the input signal multiplied by the sparse matrix specified by Matrix A.

Data Types: single | double | fixed point

Parameters

To edit block parameters interactively, use the Property Inspector. From the Simulink^® Toolstrip, on the Simulation tab, in the Prepare gallery, select Property Inspector.

Main

Matrix A — Sparse matrix

1 (default) | real scalar, vector, or matrix

Specify the sparse matrix to multiply the input vector by. Matrix A can be a real scalar, vector, or matrix.

Programmatic Use

To get the block parameter value programmatically, use the get_param function.

Parameter:	A
Values:	"1" | real scalar, vector, or matrix

Signal Attributes

Output minimum — Minimum output value for range checking

[] (default) | scalar

Specify the minimum value of the output range that the software checks.

The software uses the output minimum to perform:

• Parameter range checking (see Specify Minimum and Maximum Values for Block Parameters) for some blocks

• Simulation range checking (see Specify Signal Ranges and Enable Simulation Range Checking)

• Automatic scaling of fixed-point data types

Programmatic Use

To set the block parameter value programmatically, use the set_param function.

Parameter:	OutMin
Values:	'[]' (default) | scalar in quotes

Output maximum — Maximum output value for range checking

[] (default) | scalar

Specify the maximum value of the output range that the software checks.

The software uses the output maximum value to perform:

• Parameter range checking (see Specify Minimum and Maximum Values for Block Parameters) for some blocks

• Simulation range checking (see Specify Signal Ranges and Enable Simulation Range Checking)

• Automatic scaling of fixed-point data types

Programmatic Use

To set the block parameter value programmatically, use the set_param function.

Parameter:	OutMax
Values:	'[]' (default) | scalar in quotes

Output data type — Output data type

'Inherit: Inherit via internal rule' (default) | 'Inherit: Keep MSB' | 'Inherit: Match scaling' | 'Inherit: Inherit via back propagation' | 'Inherit: Same as first input' | 'double' | 'single' | 'half' | 'int8' | 'uint8' | 'int16' | 'uint16' | 'int32' | 'uint32' | 'int64' | 'uint64' | 'fixdt(1,16)' | 'fixdt(1,16,0)' | 'fixdt(1,16,2^0,0)' | '<data type expression>'

Choose the data type for the output. The type can be inherited, specified directly, or expressed as a data type object such as Simulink.NumericType.

Programmatic Use

To set the block parameter value programmatically, use the set_param function.

Parameter:	OutDataTypeStr
Values:	'Inherit: Inherit via internal rule' (default) | 'Inherit: Keep MSB' | 'Inherit: Match scaling' | 'Inherit: Inherit via back propagation' | 'Inherit: Same as first input' | 'double' | 'single' | 'half' | 'int8' | 'uint8' | 'int16' | 'uint16' | 'int32' | 'uint32' | 'int64' | 'uint64' | 'fixdt(1,16)' | 'fixdt(1,16,0)' | 'fixdt(1,16,2^0,0)' | '<data type expression>'

Parameter Attributes

Parameter minimum — Minimum value of sparse matrix

[] (default) | scalar

Specify the minimum value of Matrix A. The default value is [] (unspecified). The software uses this value to perform:

• Parameter range checking (see Specify Minimum and Maximum Values for Block Parameters)

• Automatic scaling of fixed-point data types

Programmatic Use

To set the block parameter value programmatically, use the set_param function.

Parameter:	ParamMin
Values:	'[]' (default) | scalar in quotes

Parameter maximum — Maximum value of sparse matrix

[] (default) | scalar

Specify the maximum value of Matrix A. The default value is [] (unspecified). The software uses this value to perform:

• Parameter range checking (see Specify Minimum and Maximum Values for Block Parameters)

• Automatic scaling of fixed-point data types

Programmatic Use

To set the block parameter value programmatically, use the set_param function.

Parameter:	ParamMax
Values:	'[]' (default) | scalar in quotes

Parameter data type — Specify the data type of the sparse matrix

'Inherit: Inherit via back propagation' (default) | 'double' | 'single' | 'half' | 'int8' | 'uint8' | 'int16' | 'uint16' | 'int32' | 'uint32' | 'int64' | 'uint64' | ' | 'fixdt(1,16)' | 'fixdt(1,16,0)' | 'fixdt(1,16,2^0,0)' | 'Enum: <class name>' | '<data type expression>'

Specify the data type of the Matrix A parameter.

While tuning a parameter with this Parameter data type setting, specify the Parameter minimum and Parameter maximum parameters. These settings tell the software about the range of values you want during the simulation and allows the software to provide a full precision data type with sufficient range to allow safe tuning within the specified range.

Programmatic Use

To set the block parameter value programmatically, use the set_param function.

Parameter:	ParamDataTypeStr
Values:	'Inherit: Inherit via back propagation' (default) | 'double' | 'single' | 'half' | 'int8' | 'uint8' | 'int16' | 'uint16' | 'int32' | 'uint32' | 'int64' | 'uint64' | 'boolean' | 'fixdt(1,16)' | 'fixdt(1,16,0)' | 'fixdt(1,16,2^0,0)' | 'Enum: <class name>' | '<data type expression>'

Extended Capabilities

HDL Code Generation

Generate VHDL, Verilog and SystemVerilog code for FPGA and ASIC designs using HDL Coder™.

HDL Coder™ provides additional configuration options that affect HDL implementation and synthesized logic.

HDL Architecture

This block has one default HDL architecture.

HDL Block Properties

By default, the following HDL block property is available.

Native Floating Point
LatencyStrategy	Specify whether to map the blocks in your design to inherit, Max, Min, or Zero for the floating-point operator. The default is inherit. See also LatencyStrategy.

Version History

Introduced in R2024b