Exercícios de FORTRAN90/95

Transcription

1 Eercícios de FORTRAN90/95 1. Estruturas de Repetição (Do Loops) 2. Estruturas codicioais (IF, CASE) 3. Arrays(1D, 2D) 4. Fuções e loops em diferetes liguages de programação. 1. Estrutura de Repetição ( Loops) Eemplo 1.1 Diferetes tipos de -Loop Método de Newto ara raiz quadrada: 1 A + 1 = ( + ) Þ 2 A (i) loop Cotrolado; repete um úmero de vezes fiado PROGRAM NEWTON REAL A! umber to be square-rooted REAL X! curret value of root INTEGER N! loop couter PRINT *, 'Eter a umber' READ *, A! iput umber to be rooted X = 1.0! iitial value N = 1, 10! fied umber of iteratios X = 0.5 * ( X + A / X )! update value (ii) loops Fleível, eecuta até que alguma codição seja satisfeita. (a) Usado IF (...) EXIT PROGRAM NEWTON REAL A!umber to be square-rooted REAL X, XOLD! curret ad previous value REAL CHANGE! chage durig oe iteratio REAL, PARAMETER :: TOLERANCE = 1.0E-6 PRINT *, 'Eter a umber' READ *, A X = 1.0 XOLD = X X = 0.5 * ( X + A / X ) CHANGE = ABS( (X - XOLD) / X ) IF ( CHANGE < TOLERANCE ) EXIT! tolerace for covergece! iput umber to be rooted! iitial value! store previous value! update value! fractioal chage! criterio for stoppig (b) Usado WHILE (...)

2 PROGRAM NEWTON REAL A REAL X, XOLD REAL CHANGE REAL, PARAMETER :: TOLERANCE = 1.0E-6 PRINT *, 'Eter a umber' READ *, A X = 1.0 CHANGE = 1.0 WHILE ( CHANGE > TOLERANCE ) XOLD = X X = 0.5 * ( X + A / X ) CHANGE = ABS( (X - XOLD) / X )! umber to be square-rooted!curret ad previous value!chage durig oe iteratio!tolerace for covergece! iput umber to be rooted! iitial value! aythig big eough to make! the first loop ru! criterio for cotiuig! store previous value! update value! fractioal chage Eemplo 1.2 Somatório de série de potêcia ep( ) = ! + 3 3! +... Note que cada termo ão é determiado por si só, mas de uma maeira mais eficiete como um múltiplo do termo previamete determiado.! = -1 ( -1)! PROGRAM POWER_SERIES REAL, EXTERNAL :: NEW_EXP! declare a fuctio to be used REAL VALUE! umber to test PRINT *, 'Eter a umber' READ *, VALUE PRINT *, 'Sum of series = ', NEW_EXP( VALUE )! our ow fuctio PRINT *, 'Actual EXP(X) = ', EXP( VALUE )! stadard fuctio STOP END PROGRAM POWER_SERIES!========================================================================= REAL FUNCTION NEW_EXP( X )! Sum a power series for ep(x) REAL X!argumet of fuctio INTEGER N!umber of a term REAL TERM!a term i the series REAL, PARAMETER :: TOLERANCE = 1.0E-07!trucatio level! First term N = 0; TERM = 1; NEW_EXP = TERM! Add successive terms util they become egligible WHILE ( ABS( TERM ) > TOLERANCE )! criterio for cotiuig N=N+1! ide of et term TERM = TERM * X / N! ew term is a multiple of last NEW_EXP = NEW_EXP + TERM END FUNCTION NEW_EXP! add to sum

3 Observação: o térmio do programa é assegurado pelo critério: term < úmero pequeo Este critério é valido desde que a presete série é covergete. Isto ão é sempre valido, logo ão é uma codição suficiete. Por eemplo, a série harmôica: å = esta série diverge, mesmo embora os termos tedem à zero. 2. Cotrole codicioal (IF, CASE) Eemplo 2.1 Comparado IF e CASE. PROGRAM EXAM INTEGER MARK CHARACTER GRADE WRITE( *, '("Eter mark (egative to ed): ")', ADVANCE = 'NO' ) READ *, MARK IF ( MARK < 0 ) STOP! stop program with a egative value IF ( MARK >= 70 ) THEN GRADE = 'A' ELSE IF ( MARK >= 60 ) THEN GRADE = 'B' ELSE IF ( MARK >= 50 ) THEN GRADE = 'C' ELSE IF ( MARK >= 40 ) THEN GRADE = 'D' ELSE IF ( MARK >= 30 ) THEN GRADE = 'E' ELSE GRADE = 'F' END IF PRINT *, 'Grade is ', GRADE END PROGRAM EXAM PROGRAM EXAM INTEGER MARK CHARACTER GRADE WRITE( *, '("Eter mark (egative to ed): ")', ADVANCE = 'NO' ) READ *, MARK IF ( MARK < 0 ) STOP! stop program with a egative value SELECT CASE ( MARK ) CASE ( 70: ) GRADE = 'A' CASE ( 60:69 ) GRADE = 'B' CASE ( 50:59 ) GRADE = 'C' CASE ( 40:49 ) GRADE = 'D' CASE ( 30:39 ) GRADE = 'E'

4 CASE ( :29 ) GRADE = 'F' END SELECT PRINT *, 'Grade is ', GRADE END PROGRAM EXAM 3. Matrizes multidimesioais - Arrays Eemplo 3.1 Ilustra operações de elemeto por elemeto com arrays PROGRAM MATRIX REAL, DIMENSION(3,3) :: A, B, C!declare size of A, B ad C! REAL A(3,3), B(3,3), C(3,3)!alterative dimesio statemet REAL PI!the umber pi INTEGER I, J!couters CHARACTER (LEN=*), PARAMETER :: FMT = '( A, 3(/, 3(1X, F8.3)), / )'! format strig for output! Basic iitialisatio of matrices by assigig all values iefficiet A(1,1) = 1.0; A(1,2) = 2.0; A(1,3) = 3.0 A(2,1) = 4.0; A(2,2) = 5.0; A(2,3) = 6.0 A(3,1) = 7.0; A(3,2) = 8.0; A(3,3) = 9.0 B(1,1) = 10.0; B(1,2) = 20.0; B(1,3) = 30.0 B(2,1) = 40.0; B(2,2) = 50.0; B(2,3) = 60.0 B(3,1) = 70.0; B(3,2) = 80.0; B(3,3) = 90.0! Alterative iitialisatio usig DATA statemets ote order DATA A / 1.0, 4.0, 7.0, 2.0, 5.0, 8.0, 3.0, 6.0, 9.0 / DATA B / 10.0, 40.0, 70.0, 20.0, 50.0, 80.0, 30.0, 60.0, 90.0 /! Alterative iitialisatio computig each elemet of A J = 1, 3 I = 1, 3 A(I,J) = (I - 1) * 3 + J! the whole-array operatio for B B = 10.0 * A! Write out matrices (usig implied loops) WRITE( *, FMT ) 'A', ( ( A(I,J), J = 1, 3 ), I = 1, 3 ) WRITE( *, FMT ) 'B', ( ( B(I,J), J = 1, 3 ), I = 1, 3 )! Matri sum C=A+B WRITE( *, FMT ) 'A+B', ( ( C(I,J), J = 1, 3 ), I = 1, 3 )! "Elemet-by-elemet" multiplicatio C=A*B WRITE( *, FMT ) 'A*B', ( ( C(I,J), J = 1, 3 ), I = 1, 3 )! "Proper" matri multiplicatio C = MATMUL( A, B ) WRITE( *, FMT ) 'MATMUL(A,B)', ( ( C(I,J), J = 1, 3 ), I = 1, 3 )! Some operatio applied to all elemets of a matri PI = 4.0 * ATAN( 1.0 ) C = SIN( B * PI / ) WRITE( *, FMT ) 'SIN(B)', ( ( C(I,J), J = 1, 3 ), I = 1, 3 ) STOP END PROGRAM MATRIX

5 Fuções e do loops em diferetes liguages de programação Cosidere a fução sumsqr ( ) = Fortra Iteger Fuctio sumsqr( ) Iteger Iteger i sumsqr = 0 Do i = 1, sumsqr = sumsqr + i * i Ed Do Ed Fuctio sumsqr! declare argumet type! declare iteral variables! iitialise sum! start of loop! add to sum! ed of loop Visual Basic Fuctio sumsqr( As Iteger) As Iteger Dim i As Iteger ' declare iteral variables sumsqr = 0 ' iitialise sum For i = 1 To ' start of loop sumsqr = sumsqr + i * i ' add to sum Net i ' ed of loop Ed Fuctio C++ it sumsqr( it ) { it i, value; // declare iteral variables value = 0; // iitialise sum for (i = 1; i <= ; i++) { // start of loop value += i * i; // add to sum } // ed of loop retur value; }