Fortran 95
Exercises
CSC – Tieteen tietotekniikan keskus Oy
CSC – IT Center for Science Ltd.
Exercise material
You can find the material (exercise examples and
exercises) from the course web site.
The example answers will be available on-line when the
course ends.
http://www.csc.fi/english/csc/courses/archive/fortran2010
2
About exercises
The course will follow closely the CSC’s guide Fortran
95/2003 (only in Finnish). See:
http://www.csc.fi/csc/julkaisut/oppaat
http://www.csc.fi/csc/julkaisut/oppaat/pdfs/fortran2003
The number of the exercise inside the parenthesis refers
to the numbering of the exercise in the Guide Fortran
95/2003.
3
Emacs editor and fortran modes
• For editing Fortran program files you can use Emacs
editor without (the option –nw) or with X-windows:
emacs –nw test.f90
emacs prog.f90
• Emacs fortran mode: On some Linux workstations it may
be necessary first load f90 lisp library: Esc X load-
library and then f90. Indentation and highlighting
works then by default.
4
Fortran compiler
The GNU Compiler Collection (GCC) Fortran 95
compiler gfortran is used. It supports also FORTRAN
77 and Fortran 90. The newest versions implement also
some or all Fortran 2003 and Fortran 2008 features. The
versions available on Linux workstations are 4.1.2 and
4.4.0
The GCC online documentation of the latest versions:
http://gcc.gnu.org/onlinedocs/
On Linux workstations the command gfortran
(version 4.1.2) and gfortran44 (version 4.4.0) can
be used directly to launch the GCC Fortran compiler.
5
Using the Gfortran compiler
Basic usage
In order to compile the file mycode1.f95 (a source code
file)
$ gfortran -c mycode1.f95
The output file will be mycode1.o (a object code file)
You can link your object files together to generate an
executable
$ gfortran -o ExecutableName mycode1.o mycode2.o mycode3.o
...
Here the executable will be called ExecutableName and the
object files were created as explained earlier. If the flag
-o ExecutableName is omitted, then the executable will be
named as a.out (on UNIX systems). 6
Using the Gfortran compiler
Basic usage ...
You can skip the separate compilation step by entering
$ gfortran -o ExecutableName mycode1.f95 mycode2.f95
mycode3.f95 ...
This will compile the source files ( mycode1.f95 mycode2.f95
mycode3.f95 ...), link and generate an executable
ExecutableName – all in one step. You could also add object
file names to this command line and they would automatically
be linked in during the link step.
Source files whose names end with .f90 or .f95 are assumed to
be free form -files. If file name ends with .f or .for then
the files are assumed to be fixed form -files (the fixed form
is an older Fortran source code style not recommended any
more).
Simple compilation
– Example program test.f95:
PROGRAM test
IMPLICIT NONE
INTEGER, PARAMETER :: result = 42
! Simple output statement...
PRINT *, ’Hello world!’
PRINT *, ’The result is ’, result
END PROGRAM test
Compilation and execution:
% gfortran test.f95 -o test
% ./test
Hello world!
The result is 42
8
Lecture#1 : Introduction
9
Exercise: Input format of the code
1. (4.1) What means the following (free format) code line
A = 0.0 ; B = 370 ! Initialization ; C = 17.0 ; D = 33.0
2. (4.2) Is the following syntactically correct Fortran 90
code?
Y = SIN(MAX(X1,X2)) * EXP(-COS(X3)**I) - TAN(AT&
& AN(X4))
3. (4.3) Is the following declaration correct?
REAL :: real; REAL :: price
10
1. (4.5) Are the following Fortran statements written
correctly?
character_string = ’Awake in the morning,
& asleep in the evening.’
x = 0.4-6
answer = ’True & false’
low-limit = 0.0005E10
y = E6
11
Exercise: Declaration of variables
1. (5.1) Are the following declarations legal in Fortran:
DOUBLE :: x
CHARACTER(LEN=*), PARAMETER :: "Name"
REAL :: pi = 22/7
REAL :: x = 2., y = -3
REAL :: pii = 22.0/7.0
REAL x = 1.0
2. (5.2) Declare the constant k, whose value is 0.75.
12
1. (5.5) Find out, what are the ranges of the values of integer and real
numbers, which the Fortran compiler you use can handle. Hint:
study the values which the following subprogram calls return
SELECTED_REAL_KIND
SELECTED_INT_KIND
e.g., by the following way:
PROGRAM kindtest
IMPLICIT NONE
INTEGER :: i, ikind
DO i = 1, 24
ikind = SELECTED_INT_KIND(i)
WRITE (*,*) ’i = ’,i, ’ kind = ’,ikind
END DO
END PROGRAM kindtest
Please, note that the values of the kind parameter depend on the
Fortran compiler used!
13
1. (5.6) Print the value of the statement
using at least the precision of 6 decimals. Use the
function calls LOG and SQRT. In addition, print the
value of the statement using the highest precision of
real numbers which the compiler knows.
14
Exercise: Control structures
1. (7.3) What are the iteration counts of the following DO loops, the
values of loop variable i inside the loop, and the value of the loop
variable after the DO construct?
DO i = 1, 5
DO i = 5, 0, -1
DO i = 10, 1, -2
DO i = 0, 30, 7
DO i = 3, 2, 1
1. (7.2) Write a SELECT CASE structure, which does different
operations when an integer variable is negative, it is zero, or it is
one of the prime numbers (3, 5, 7, 11, 13). In other cases nothing is
done.
15
1. (7.6) Print the transformation table for converting inches
to millimeters (1" = 25.4mm). Start from the value 0.0
and print values at half inch intervals up to 12 inches.
16
