Hands-on

These are some exercises for you to try.

Fortran code can be writing using any text editor. A Fortran file, e.g., my_program.f90 can be compiled using

$ gfortran  -Wall  -g  -o my_program.exe  my_program.f90

If there are not compilation or linking errors, you can run the application as

$ ./my_program.exe

Note:

• on VSC infrastructure, load the foss module,
• on your Linux laptop, make sure you installed gfortran using apt, yum or whichever package manager is appropriate for your system.
• on your Windows laptop, you can likely use either Cygwin, or, on Windows 10, use the Linux Subsystem for Windows (LWS).

Hands-on session 1: control flow statements, data types, functions

1. Write a Fortran program that writes the value of the factorial to the screen for the number 0 through 10. The factorial of n is defined as n! = 1*2*3*…*(n - 1)*n.

2. Write a Fortran program that checks whether the numbers 1 though 40 are prime numbers, writing the result of each test to the screen. A number is prime when it can only be divided by 1 and itself, and 1 is not prime.

3. Write a Fortran program that for all integers n between 1 and 20 and all integers p between 1 and 20 prints to screen the greatest common divisor of n and p. For instance, the greatest common divisor of 12 and 15 is 3.

Hands-on session 2: Arrays and procedures

1. Initialize two 2D arrays with all elements equal to double precision 1.0. The first array is 1000 by 2000, the second 2000 by 300. Write two versions of this program,
   • one that implements the dot-product of two matrices as three nested iterations, i.e., C(i, j) = A(i, 1)*B(1, j) + A(i, 2)*B(2, j) + ... + A(i, 2000)*B(2000, j) for all i between 1 and 1000, and all j between 1 and 3000.
   • one that used Fortran’s matmul function.

   Compare the performance of these two programs.

2. Initialize three 2D 20000 by 3000 arrays with all elements equal to double precision 1.0. Write three applications to compute D = 5.0*A*B + C/1.5 + 8.0,
   • the first uses the given array expression,
   • the second uses FORALL (can you use this statement here?),
   • the third has explicit DO loops over the array indices.

   Compare the performance of these three programs.

3. If you didn’t already do so in the first hands-on, implement functions factorial and is_prime respectively, with the obvious semantics.

4. Write a Fortran program that computes and prints the Fibonacci numbers for 0 through 20 to the screen. The Fibonacci number fib(n) for n is defined as fib(n - 1) + fib(n - 2), and fib(0) = 1, fib(1) = 1.

   As an extra challenge, implment a non-recursive Fibonacci function, and compare the performance of the recursive and the non-recursive version.

5. Write a procedure that modifies a 2D array containing duoble precision floating point values passed as an argument, and modifying it in-place. The elements of the 2D array should be set to 0.0 if the element’s absolute value is less than a given real number that is given as the second argument to the function. Pay attention to the intent of the arguments.

Hands-on session 3: user-defined types, modules and object-oriented programming

1. Write a module that defines a user-defined type to represent descriptive statistics information. It should have the fields sum, sum2 and n representing the sum of the data, the sum of the squares of the data, and the number of data items respectively. The module should also define a procedure to initialize a variable of the user defined type, that adds a data value, one that takes a variable of this type as an argument and computes the mean value, and another one that computes the standard deviation.

2. Write a module that defines a class to maintain statistics over a sliding window. You can add as many data values to such an object, but it will compute descriptive statistics (mean, standard deviation) over a window of a size you specify when the object is created.

Hands-on session 4: file I/O and interacting with the environment

1. Write a Fortran program that reads a text file that contains floating point numbers, and that computes descriptive statistics (mean, standard deviation) of that data.

2. Write a Fortran program that reads a text file that has the following format:

   • the first line contains two integer numbers, representing the dimensions of a 2D array, the first is the number of rows, the second the number of columns.
   • the following lines each represent a row of the 2D array.

   The program multiplies each element of the array by a number that is given as a second command line argument, and writes the resulting matrix to a file, the name of which is given as a third command line argument.

   You find an example of such a file in the repository.