mrm.execution.checkModel

Check Modelscape model validity

Since R2023a

Syntax

result = mrm.execution.checkModel(model,inputData)

result = mrm.execution.checkModel(model,inputData,parameterData)

Description

Add-On Required: This feature requires the Modelscape for MATLAB add-on.

result = mrm.execution.checkModel(model,inputData) checks the validity of a Modelscape™ model by creating several test methods. The function determines whether the Modelscape model has the right shape for the input of the original model. The function returns the result of the tests in result. This function requires the Modelscape for MATLAB^® support package.

result = mrm.execution.checkModel(model,inputData,parameterData) specifies additional parameters for the model.

Examples

Check Validity of Modelscape Implementation of Generalized Linear Model (GLM)

Create a Modelscape model as a subclass of the mrm.execution.Model class.

The model takes the inputs X1, and X2, as well as an optional parameter intercept. The function computes the output Y as the value of exp(intercept + alpha1*X1 + alpha2*X2).

The Modelscape model must also implement three methods for specifying the inputs and the outputs of the model: getInputs, getParameters, and getOutputs. For more details, see the Model Implementation for Modelscape Deploy

classdef glmModel < mrm.execution.Model
    %   implement simple GLM with two inputs X1 and X2 that computes
    %   Y = exp(intercept + alpha1*X1 + alpha2*X2)
    properties
        alpha1 = 1
        alpha2 = 2
    end


    methods
        
         function parameters = getInputs(~)
            doubleDatatype = struct( ...
                "name", "double");
            parameters = struct( ...
                "name", {"X1","X2"}, ...
                "dataType", {doubleDatatype, doubleDatatype});
        end

        function parameters = getParameters(~)
            doubleDatatype = struct( ...
                "name", "double");
            parameters = struct( ...
                "name", {"intercept"}, ...
                "dataType", {doubleDatatype});
        end

        function parameters = getOutputs(~)
            doubleDatatype = struct( ...
                "name", "double");
            parameters = struct( ...
                "name", {"Y"}, ...
                "dataType", {doubleDatatype});
        end

        function [outputs,diagnostics,batchDiagnostics] = evaluate(this,inputs,parameters)
            outputs = table( ...
                exp(parameters.intercept +  this.alpha1*inputs.X1 + this.alpha2*inputs.X2), ...
                'VariableNames', {'Y'}, ...
                'RowNames', inputs.Properties.RowNames);
            rawDiagnostics = [inputs.Properties.RowNames repmat({struct()},numel(inputs.Properties.RowNames),1)]';
            diagnostics = struct(rawDiagnostics{:});
            batchDiagnostics = struct();
        end

    end
end

Validate the custom model.

result = mrm.execution.checkModel(model, inputData, parameterData);

Input Arguments

model — Modelscape model

subclass of mrm.execution.Model

Modelscape model, specified as a subclass of the mrm.execution.Model class. This model must also implement three methods for specifying the inputs and the outputs of the model: getInputs, getParameters, and getOutputs. For more details, see Model Implementation for Modelscape Deploy.

inputData — Original model inputs

table

Original model inputs, specified as a table.

Data Types: table

parameterData — Original model parameters

structure

Original model parameters, specified as a structure.

Data Types: struct

Output Arguments

result — Test results

matlab.unittest.TestResult class

Test results, returned as a matlab.unittest.TestResult class. The results include information describing whether the test pass, fail, or run to completion, as well as the duration of each test.

Version History

Introduced in R2023a