Command Line Arguments

This tutorial is automatically generated from TestCommandLineArgumentsTutorial.hpp at revision c91d6a04c957. Note that the code is given in full at the bottom of the page.

#include <cxxtest/TestSuite.h>
#include <iostream>
#include <iomanip>

Most Chaste code uses PETSc to solve linear algebra problems. This involves starting PETSc at the beginning of a test-suite and closing it at the end. (If you never run code in parallel then it is safe to replace PetscSetupAndFinalize.hpp with FakePetscSetup.hpp)

#include "PetscSetupAndFinalize.hpp"
#include "CommandLineArguments.hpp"

/**
 * @file
 *
 * This is a tutorial for providing command line arguements to a test from a bash script.
 * The example bash script also includes a for loop as an example of how such input could
 * be used to run multiple tests with varying parameter inputs.
 */

 // NOTE: These tests will run if executed from terminal directly with their default varaible values. However, these
 // tutorial tests are intended to be executed through the use of bash/shell scripts (.sh files).
class TestCommandLineArgumentsTutorial : public CxxTest::TestSuite
{
public:

To script the running of this test with lots of different arguments, copy and paste the following code into a bash script (.sh) file. This bash script file will need to be saved in your chaste build folder.

#!/bin/bash

# This bash script accompanies that TestCommandLineArgumentsTutorial. This script will execute the test TestCommandLineArgumentsTutorial
# multiple times with several command line arguments as test variables.
# Here we will declare some values we wish to later pass to a for loop.

# Here we will declare some values we wish to later pass to a for loop.
N=2
L=3
M=4

# Here we set up a simple for loop over variables i,j and k based on the values of N,L and M.
for ((i = 0; i <= N; i += 1)); do
for ((j = 1; j <= L; j += 1)); do
for ((k = 2; k <= M; k += 1)); do
# Each loop runs an instance of the TestCommandLineArgumentsTutorial with opt1,opt2 and opt3 taking on the
# values of i,j and k resepctivley.
./global/test/TestCommandLineArgumentsTutorial -opt1 $i -opt2 $j -opt3 $k &
done
done
done

    void TestCommandLineDefaultTutorial()
    {
        // First, we set up some variables. These variables should later be assigned our command line argument values.
        // However we will give them default values to ensure that this test can always run.
        unsigned outp1 = 1;
        unsigned outp2 = 2;
        unsigned outp3 = 3;

        // Here an if statement is utilised to check that we have passed some command line arguments to our test.
        if (CommandLineArguments::Instance()->OptionExists("-opt1") && CommandLineArguments::Instance()->OptionExists("-opt2") && CommandLineArguments::Instance()->OptionExists("-opt3"))
        {
           // If command line arguments have been passed to our test, we utilise CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption()
           // to take in our command line arguements.
           outp1 = CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption("-opt1");
           outp2 = CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption("-opt2");
           outp3 = CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption("-opt3");
        }

        unsigned sum = outp1 + outp2 + outp3;

        std::cout << "When we add "<< outp1 << " ,"<<  outp2 << " and "<< outp3 <<" we get " << sum <<"\n";
    }

In addition to using unsigned integers as command line arguments we can also pass in both doubles and a vector of arguments. This could be very useful for several reasons, such as having a large list of parameters one may wish to pass into a test.

To script the running of this test with lots of different arguments, copy and paste the following code into a bash script (.sh) file.

# Here we will declare some values we wish to later pass to a for loop.
# As bash script cannot directly handle double arithmetic we will First
# declare these as larger variables to be handled by another programme later.
N=2000
L=3000
M=4000

# Here we set up a simple for loop over variables i,j and k based on the values of N,L and M.
for ((i = 1; i <= N; i += 1000)); do
for ((j = 1001; j <= L; j += 1000)); do
for ((k = 2001; k <= M; k += 1000)); do
# As bash cannot directly handle double floating point arithmetic we will utilise
# awk to divide our varaibles and convert them to doubles.
idouble=$(awk "BEGIN {printf \"%.2f\",$i/120}")
jdouble=$(awk "BEGIN {printf \"%.2f\",$j/150}")
kdouble=$(awk "BEGIN {printf \"%.2f\",$k/180}")

# Each loop runs an instance of the TestCommandLineArgumentsTutorial with a vector
# containing the double corrected version of our varaibles.
./global/test/TestCommandLineArgumentsTutorial --my-vector-of-arguments $idouble $jdouble $kdouble &
done
done
done

    void TestCommandLineDoubleTutorial()
    {
        // Here an if statement is utilised to check that we have passed some command line arguments to our test.
        // To ensure the test can still run in the absence of command line arguments we will initialise the vector
        // with some default values if no command line option exists.
        if (CommandLineArguments::Instance()->OptionExists("--my-vector-of-arguments"))
        {
           // First, we need to setup our vector based on the vector passed in as our command line arguement.
           std::vector<double> vector_of_doubles = CommandLineArguments::Instance()->GetDoublesCorrespondingToOption("--my-vector-of-arguments");
           double sum = 0.0;
           // Now summing over our vector.
           for(double i = 0; i < vector_of_doubles.size(); i++){
            sum += vector_of_doubles[i];
           }
           // We will finally cout our result for the summed vector components.
           // Here we have chosen to set our precision to 4 to ensure the correct number of significant figures.
           std::cout << std::setprecision(5) << "When we add "<< vector_of_doubles[0] << ", " << vector_of_doubles[1] << " and " << vector_of_doubles[2] << " together from our vector we get " << sum  <<std::endl;

        }
    }
};

Full code

#include <cxxtest/TestSuite.h>
#include <iostream>
#include <iomanip>
#include "PetscSetupAndFinalize.hpp"
#include "CommandLineArguments.hpp"

/**
 * @file
 *
 * This is a tutorial for providing command line arguements to a test from a bash script.
 * The example bash script also includes a for loop as an example of how such input could
 * be used to run multiple tests with varying parameter inputs.
 */

 // NOTE: These tests will run if executed from terminal directly with their default varaible values. However, these
 // tutorial tests are intended to be executed through the use of bash/shell scripts (.sh files).
class TestCommandLineArgumentsTutorial : public CxxTest::TestSuite
{
public:

    void TestCommandLineDefaultTutorial()
    {
        // First, we set up some variables. These variables should later be assigned our command line argument values.
        // However we will give them default values to ensure that this test can always run.
        unsigned outp1 = 1;
        unsigned outp2 = 2;
        unsigned outp3 = 3;

        // Here an if statement is utilised to check that we have passed some command line arguments to our test.
        if (CommandLineArguments::Instance()->OptionExists("-opt1") && CommandLineArguments::Instance()->OptionExists("-opt2") && CommandLineArguments::Instance()->OptionExists("-opt3"))
        {
           // If command line arguments have been passed to our test, we utilise CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption()
           // to take in our command line arguements.
           outp1 = CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption("-opt1");
           outp2 = CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption("-opt2");
           outp3 = CommandLineArguments::Instance()->GetUnsignedCorrespondingToOption("-opt3");
        }

        unsigned sum = outp1 + outp2 + outp3;

        std::cout << "When we add "<< outp1 << " ,"<<  outp2 << " and "<< outp3 <<" we get " << sum <<"\n";
    }

    void TestCommandLineDoubleTutorial()
    {
        // Here an if statement is utilised to check that we have passed some command line arguments to our test.
        // To ensure the test can still run in the absence of command line arguments we will initialise the vector
        // with some default values if no command line option exists.
        if (CommandLineArguments::Instance()->OptionExists("--my-vector-of-arguments"))
        {
           // First, we need to setup our vector based on the vector passed in as our command line arguement.
           std::vector<double> vector_of_doubles = CommandLineArguments::Instance()->GetDoublesCorrespondingToOption("--my-vector-of-arguments");
           double sum = 0.0;
           // Now summing over our vector.
           for(double i = 0; i < vector_of_doubles.size(); i++){
            sum += vector_of_doubles[i];
           }
           // We will finally cout our result for the summed vector components.
           // Here we have chosen to set our precision to 4 to ensure the correct number of significant figures.
           std::cout << std::setprecision(5) << "When we add "<< vector_of_doubles[0] << ", " << vector_of_doubles[1] << " and " << vector_of_doubles[2] << " together from our vector we get " << sum  <<std::endl;

        }
    }
};

Combining consecutive cell-based simulations

Creating And Using A New Cell Based Simulation Modifier

Command Line Arguments

Full code#

Full code