Cramer's rule

From Rosetta Code
Task
Cramer's rule
You are encouraged to solve this task according to the task description, using any language you may know.
In linear algebra, Cramer's rule is an explicit formula for the solution of a system of linear equations with as many equations as unknowns, valid whenever the system has a unique solution. It expresses the solution in terms of the determinants of the (square) coefficient matrix and of matrices obtained from it by replacing one column by the vector of right hand sides of the equations.


Given

which in matrix format is

Then the values of and can be found as follows:


Task

Given the following system of equations:

solve for , , and , using Cramer's rule.

ALGOL 68[edit]

Works with: ALGOL 68G version Any - tested with release 2.8.3.win32

Uses the non-standard DET operator available in Algol 68G.

# returns the solution of a.x = b via Cramer's rule                    #
# this is for REAL arrays, could define additional operators #
# for INT, COMPL, etc. #
PRIO CRAMER = 1;
OP CRAMER = ( [,]REAL a, []REAL b )[]REAL:
IF 1 UPB a /= 2 UPB a
OR 1 LWB a /= 2 LWB a
OR 1 UPB a /= UPB b
THEN
# the array sizes and bounds do not match #
print( ( "Invaid parameters to CRAMER", newline ) );
stop
ELIF REAL deta = DET a;
det a = 0
THEN
# a is singular #
print( ( "Singular matrix for CRAMER", newline ) );
stop
ELSE
# the arrays have matching bounds #
[ LWB b : UPB b ]REAL result;
FOR col FROM LWB b TO UPB b DO
# form a matrix from a with the col'th column replaced by b #
[ 1 LWB a : 1 UPB a, 2 LWB a : 2 UPB a ]REAL m := a;
m[ : , col ] := b[ : AT 1 ];
# col'th result elemet as per Cramer's rule #
result[ col ] := DET m / det a
OD;
result
FI; # CRAMER #
 
# test CRAMER using the matrix and column vector specified in the task #
[,]REAL a = ( ( 2, -1, 5, 1 )
, ( 3, 2, 2, -6 )
, ( 1, 3, 3, -1 )
, ( 5, -2, -3, 3 )
);
[]REAL b = ( -3
, -32
, -47
, 49
);
[]REAL solution = a CRAMER b;
FOR c FROM LWB solution TO UPB solution DO
print( ( " ", fixed( solution[ c ], -8, 4 ) ) )
OD;
print( ( newline ) )
 
Output:
   2.0000 -12.0000  -4.0000   1.0000

C[edit]

#include <math.h>
#include <stdio.h>
#include <stdlib.h>
 
typedef struct {
int n;
double **elems;
} SquareMatrix;
 
SquareMatrix init_square_matrix(int n, double elems[n][n]) {
SquareMatrix A = {
.n = n,
.elems = malloc(n * sizeof(double *))
};
for(int i = 0; i < n; ++i) {
A.elems[i] = malloc(n * sizeof(double));
for(int j = 0; j < n; ++j)
A.elems[i][j] = elems[i][j];
}
 
return A;
}
 
SquareMatrix copy_square_matrix(SquareMatrix src) {
SquareMatrix dest;
dest.n = src.n;
dest.elems = malloc(dest.n * sizeof(double *));
for(int i = 0; i < dest.n; ++i) {
dest.elems[i] = malloc(dest.n * sizeof(double));
for(int j = 0; j < dest.n; ++j)
dest.elems[i][j] = src.elems[i][j];
}
 
return dest;
}
 
double det(SquareMatrix A) {
double det = 1;
 
for(int j = 0; j < A.n; ++j) {
int i_max = j;
for(int i = j; i < A.n; ++i)
if(A.elems[i][j] > A.elems[i_max][j])
i_max = i;
 
if(i_max != j) {
for(int k = 0; k < A.n; ++k) {
double tmp = A.elems[i_max][k];
A.elems[i_max][k] = A.elems[j][k];
A.elems[j][k] = tmp;
}
 
det *= -1;
}
 
if(abs(A.elems[j][j]) < 1e-12) {
puts("Singular matrix!");
return NAN;
}
 
for(int i = j + 1; i < A.n; ++i) {
double mult = -A.elems[i][j] / A.elems[j][j];
for(int k = 0; k < A.n; ++k)
A.elems[i][k] += mult * A.elems[j][k];
}
}
 
for(int i = 0; i < A.n; ++i)
det *= A.elems[i][i];
 
return det;
}
 
void deinit_square_matrix(SquareMatrix A) {
for(int i = 0; i < A.n; ++i)
free(A.elems[i]);
free(A.elems);
}
 
double cramer_solve(SquareMatrix A, double det_A, double *b, int var) {
SquareMatrix tmp = copy_square_matrix(A);
for(int i = 0; i < tmp.n; ++i)
tmp.elems[i][var] = b[i];
 
double det_tmp = det(tmp);
deinit_square_matrix(tmp);
 
return det_tmp / det_A;
}
 
int main(int argc, char **argv) {
#define N 4
double elems[N][N] = {
{ 2, -1, 5, 1},
{ 3, 2, 2, -6},
{ 1, 3, 3, -1},
{ 5, -2, -3, 3}
};
SquareMatrix A = init_square_matrix(N, elems);
 
SquareMatrix tmp = copy_square_matrix(A);
int det_A = det(tmp);
deinit_square_matrix(tmp);
 
double b[] = {-3, -32, -47, 49};
 
for(int i = 0; i < N; ++i)
printf("%7.3lf\n", cramer_solve(A, det_A, b, i));
 
deinit_square_matrix(A);
return EXIT_SUCCESS;
}
Output:
  2.000
-12.000
 -4.000
  1.000

Common Lisp[edit]

(defun minor (m col)
(loop with dim = (1- (array-dimension m 0))
with result = (make-array (list dim dim))
for i from 1 to dim
for r = (1- i)
do (loop with c = 0
for j to dim
when (/= j col)
do (setf (aref result r c) (aref m i j))
(incf c))
finally (return result)))
 
(defun det (m)
(assert (= (array-rank m) 2))
(assert (= (array-dimension m 0) (array-dimension m 1)))
(let ((dim (array-dimension m 0)))
(if (= dim 1)
(aref m 0 0)
(loop for col below dim
for sign = 1 then (- sign)
sum (* sign (aref m 0 col) (det (minor m col)))))))
 
(defun replace-column (m col values)
(let* ((dim (array-dimension m 0))
(result (make-array (list dim dim))))
(dotimes (r dim result)
(dotimes (c dim)
(setf (aref result r c)
(if (= c col) (aref values r) (aref m r c)))))))
 
(defun solve (m v)
(loop with dim = (array-dimension m 0)
with det = (det m)
for col below dim
collect (/ (det (replace-column m col v)) det)))
 
(solve #2A((2 -1 5 1)
(3 2 2 -6)
(1 3 3 -1)
(5 -2 -3 3))
#(-3 -32 -47 49))
Output:
(2 -12 -4 1)

EchoLisp[edit]

 
(lib 'matrix)
(string-delimiter "")
(define (cramer A B (X)) ;; --> vector X
(define(determinant A))
(for/vector [(i (matrix-col-num A))]
(set! X (matrix-set-col! (array-copy A) i B))
(// (determinant X))))
 
(define (task)
(define A (list->array
'( 2 -1 5 1 3 2 2 -6 1 3 3 -1 5 -2 -3 3) 4 4))
(define B #(-3 -32 -47 49))
(writeln "Solving A * X = B")
(array-print A)
(writeln "B = " B)
(writeln "X = " (cramer A B)))
 
Output:
(task)
Solving A * X = B    
  2   -1   5    1  
  3   2    2    -6 
  1   3    3    -1 
  5   -2   -3   3  
B =      #( -3 -32 -47 49)    
X =      #( 2 -12 -4 1)    

Fortran[edit]

In Numerical Methods That Work (Usually), in the section What not to compute, F. S. Acton remarks "...perhaps we should be glad he didn't resort to Cramer's rule (still taught as the practical method in some high schools) and solve his equations as the ratios of determinants - a process that requires labor proportional to N! if done in the schoolboy manner. The contrast with N3 can be startling!" And further on, "Having hinted darkly at my computational fundamentalism, it is probably time to commit to a public heresy by denouncing recursive calculations. I have never seen a numerical problem arising from the physical world that was best calculated by a recursive subroutine..."

Since this problem requires use of Cramer's rule, one might as well be hung for a sheep instead of a lamb, so the traditions of Old Fortran and heavy computation will be ignored and the fearsome RECURSIVE specification employed so that the determinants will be calculated recursively, all the way down to N = 1 even though the N = 2 case is easy. This requires F90 and later. Similarly, the MODULE protocol will be employed, even though there is no significant context to share. The alternative method for calculating a determinant involves generating permutations, a tiresome process.

Array passing via the modern arrangements of F90 is a source of novel difficulty to set against the slight convenience of not having to pass an additional parameter, N. Explicitly, at least. There are "secret" additional parameters when an array is being passed in the modern way, which are referred to by the new SIZE function. Anyway, for an order N square matrix, the array must be declared A(N,N), and specifically not something like A(100,100) with usage only of elements up to N = 7, say, because the locations in storage of elements in use would be quite different from those used by an array declared A(7,7). This means that the array must be re-declared for each different size usage, a tiresome and error-inviting task. One-dimensional arrays do not have this problem, but they do have to be "long enough" so B and X might as well be included. This also means that the auxiliary matrices needed within the routines have to be made the right size, and fortunately they can be declared in a way that requests this without the blather of ALLOCATE, this being a protocol introduced by Algol in the 1960s. Unfortunately, there is no scheme such as in pl/i to declare AUX "like" A, so some grotesquery results. And in the case of function DET which needs an array of order N - 1, when its recursion bottoms out with N = 1 it will have declared MINOR(0,0), a rather odd situation that fortunately evokes no complaint, and a test run in which its "value" was written out by WRITE (6,*) MINOR produced a blank line: no complaint there either, presumably because zero elements were being sent forth and so there was no improper access of ... nothing.

With matrices, there is a problem all the way from the start in 1958. Everyone agrees that a matrix should be indexed as A(row,column) and that when written out, rows should run down the page and columns across. This is unexceptional and the F90 function MATMUL follows this usage. However, Fortran has always stored its array elements in what is called "column major" order, which is to say that element A(1,1) is followed by element A(2,1) in storage, not A(1,2). Thus, if an array is written (or read) by something like WRITE (6,*) A, consecutive elements, written along a line, will be A(1,1), A(2,1), A(3,1), ... So, subroutine SHOWMATRIX is employed to write the matrix out in the desired form, and to read the values into the array, an explicit loop is used to place them where expected rather than just READ(INF,*) A

Similarly, if instead a DATA statement were used to initialise the array for the example problem, and it looked something like
      DATA A/2, -1,  5,  1
1 3, 2, 2, -6
2 1, 3, 3, -1
3 5, -2, -3, 3/
(ignoring integer/floating-point issues) thus corresponding to the layout of the example problem, there would need to be a statement A = TRANSPOSE(A) to obtain the required order. I have never seen an explanation of why this choice was made for Fortran.
      MODULE BAD IDEA	!Employ Cramer's rule for solving A.x = b...
INTEGER MSG !Might as well be here.
CONTAINS !The details.
SUBROUTINE SHOWMATRIX(A) !With nice vertical bars.
DOUBLE PRECISION A(:,:) !The matrix.
INTEGER R,N !Assistants.
N = SIZE(A, DIM = 1) !Instead of passing an explicit parameter.
DO R = 1,N !Work down the rows.
WRITE (MSG,1) A(R,:) !Fling forth a row at a go.
1 FORMAT ("|",<N>F12.3,"|") !Bounded by bars.
END DO !On to the next row.
END SUBROUTINE SHOWMATRIX !Furrytran's default order is the transpose.
 
RECURSIVE DOUBLE PRECISION FUNCTION DET(A) !Determine the determinant.
DOUBLE PRECISION A(:,:) !The square matrix, order N.
DOUBLE PRECISION MINOR(SIZE(A,DIM=1) - 1,SIZE(A,DIM=1) - 1) !Order N - 1.
DOUBLE PRECISION D !A waystation.
INTEGER C,N !Steppers.
N = SIZE(A, DIM = 1) !Suplied via secret parameters.
IF (N .LE. 0) THEN !This really ought not happen.
STOP "DET: null array!" !But I'm endlessly suspicious.
ELSE IF (N .NE. SIZE(A, DIM = 2)) THEN !And I'd rather have a decent message
STOP "DET: not a square array!" !In place of a crashed run.
ELSE IF (N .EQ. 1) THEN !Alright, now get on with it.
DET = A(1,1) !This is really easy.
ELSE !But otherwise,
D = 0 !Here we go.
DO C = 1,N !Step along the columns of the first row.
CALL FILLMINOR(C) !Produce the auxiliary array for each column.
IF (MOD(C,2) .EQ. 0) THEN !Odd or even case?
D = D - A(1,C)*DET(MINOR) !Even: subtract.
ELSE !Otherwise,
D = D + A(1,C)*DET(MINOR) !Odd: add.
END IF !So much for that term.
END DO !On to the next.
DET = D !Declare the result.
END IF !So much for that.
CONTAINS !An assistant.
SUBROUTINE FILLMINOR(CC) !Corresponding to A(1,CC).
INTEGER CC !The column being omitted.
INTEGER R !A stepper.
DO R = 2,N !Ignoring the first row,
MINOR(R - 1,1:CC - 1) = A(R,1:CC - 1) !Copy columns 1 to CC - 1. There may be none.
MINOR(R - 1,CC:) = A(R,CC + 1:) !And from CC + 1 to N. There may be none.
END DO !On to the next row.
END SUBROUTINE FILLMINOR !Divide and conquer.
END FUNCTION DET !Rather than mess with permutations.
 
SUBROUTINE CRAMER(A,X,B) !Solve A.x = b, where A is a matrix...
Careful! The array must be A(N,N), and not say A(100,100) of which only up to N = 6 are in use.
DOUBLE PRECISION A(:,:) !A square matrix. I hope.
DOUBLE PRECISION X(:),B(:) !Column matrices look rather like 1-D arrays.
DOUBLE PRECISION AUX(SIZE(A,DIM=1),SIZE(A,DIM=1)) !Can't say "LIKE A", as in pl/i, alas.
DOUBLE PRECISION D !To be calculated once.
INTEGER N !The order of the square matrix. I hope.
INTEGER C !A stepper.
N = SIZE(A, DIM = 1) !Alright, what's the order of battle?
D = DET(A) !Once only.
IF (D.EQ.0) STOP "Cramer: zero determinant!" !Surely, this won't happen...
AUX = A !Prepare the assistant.
DO C = 1,N !Step across the columns.
IF (C.GT.1) AUX(1:N,C - 1) = A(1:N,C - 1) !Repair previous damage.
AUX(1:N,C) = B(1:N) !Place the current damage.
X(C) = DET(AUX)/D !The result!
END DO !On to the next column.
END SUBROUTINE CRAMER !This looks really easy!
END MODULE BAD IDEA !But actually, it is a bad idea for N > 2.
 
PROGRAM TEST !Try it and see.
USE BAD IDEA !Just so.
DOUBLE PRECISION, ALLOCATABLE ::A(:,:), B(:), X(:) !Custom work areas.
INTEGER N,R !Assistants..
INTEGER INF !An I/O unit.
 
MSG = 6 !Output.
INF = 10 !For reading test data.
OPEN (INF,FILE="Test.dat",STATUS="OLD",ACTION="READ") !As in this file..
 
Chew into the next problem.
10 IF (ALLOCATED(A)) DEALLOCATE(A) !First,
IF (ALLOCATED(B)) DEALLOCATE(B) !Get rid of
IF (ALLOCATED(X)) DEALLOCATE(X) !The hired help.
READ (INF,*,END = 100) N !Since there is a new order.
IF (N.LE.0) GO TO 100 !Perhaps a final order.
WRITE (MSG,11) N !Othewise, announce prior to acting.
11 FORMAT ("Order ",I0," matrix A, as follows...") !In case something goes wrong.
ALLOCATE(A(N,N)) !For instance,
ALLOCATE(B(N)) !Out of memory.
ALLOCATE(X(N)) !But otherwise, a tailored fit.
DO R = 1,N !Now read in the values for the matrix.
READ(INF,*,END=667,ERR=665) A(R,:),B(R) !One row of A at a go, followed by B's value.
END DO !In free format.
CALL SHOWMATRIX(A) !Show what we have managed to obtain.
WRITE (MSG,12) "In the scheme A.x = b, b = ",B !In case something goes wrong.
12 FORMAT (A,<N>F12.6) !How many would be too many?
CALL CRAMER(A,X,B) !The deed!
WRITE (MSG,12) " Via Cramer's rule, x = ",X !The result!
GO TO 10 !And try for another test problem.
 
Complaints.
665 WRITE (MSG,666) "Format error",R !I know where I came from.
666 FORMAT (A," while reading row ",I0,"!") !So I can refer to R.
GO TO 100 !So much for that.
667 WRITE (MSG,666) "End-of-file", R !Some hint as to where.
 
Closedown.
100 WRITE (6,*) "That was interesting." !Quite.
END !Open files are closed, allocated memory is released.

Oddly, the Compaq Visual Fortran F90/95 compiler is confused by the usage "BAD IDEA" instead of "BADIDEA" - spaces are not normally relevant in Fortran source. Anyway, file Test.dat has been filled with the numbers of the example, as follows:

4           /The order, for A.x = b.
2  -1   5   1,  -3    /First row of A, b
3   2   2  -6, -32    /Second row...
1   3   3  -1, -47     third row.
5  -2  -3   3,  49    /Last row.

Fortran's free-form allows a comma, a tab, and spaces between numbers, and regards the / as starting a comment, but, because each row is read separately, once the required five (N + 1) values have been read, no further scanning of the line takes place and the next READ statement will start with a new line of input. So the / isn't needed for the third row, as shown. Omitted values lead to confusion as the input process would read additional lines to fill the required count and everything gets out of step. Echoing input very soon after it is obtained is helpful in making sense of such mistakes.

For more practical use it would probably be better to constrain the freedom somewhat, perhaps requiring that all the N + 1 values for a row appear on one input record. In such a case, the record could first be read into a text variable (from which the data would be read) so that if a problem arises the text could be printed as a part of the error message. But, this requires guessing a suitably large length for the text variable so as to accommodate the longest possible input line.

Output:

Order 4 matrix A, as follows...
|       2.000      -1.000       5.000       1.000|
|       3.000       2.000       2.000      -6.000|
|       1.000       3.000       3.000      -1.000|
|       5.000      -2.000      -3.000       3.000|
In the scheme A.x = b, b =    -3.000000  -32.000000  -47.000000   49.000000
    Via Cramer's rule, x =     2.000000  -12.000000   -4.000000    1.000000
 That was interesting.

And at this point I suddenly noticed that the habits of Old Fortran are not so easily suppressed: all calculations are done with double precision. Curiously enough, for the specific example data, the same results are obtained if all variables are integer.

Haskell[edit]

Version 1[edit]

import Data.Matrix
 
solveCramer :: (Ord a, Fractional a) => Matrix a -> Matrix a -> Maybe [a]
solveCramer a y
| da == 0 = Nothing
| otherwise = Just $ map (\i -> d i / da) [1..n]
where da = detLU a
d i = detLU $ submatrix 1 n 1 n $ switchCols i (n+1) ay
ay = a <|> y
n = ncols a
 
task = solveCramer a y
where a = fromLists [[2,-1, 5, 1]
,[3, 2, 2,-6]
,[1, 3, 3,-1]
,[5,-2,-3, 3]]
y = fromLists [[-3], [-32], [-47], [49]]
Output:
λ> task
Just [2.0000000000000004,-11.999999999999998,-4.0,0.9999999999999999]

Version 2[edit]

We use Rational numbers for having more precision. a % b is the rational a / b.

 
s_permutations :: [a] -> [([a], Int)]
s_permutations = flip zip (cycle [1, -1]) . (foldl aux [[]])
where aux items x = do
(f,item) <- zip (cycle [reverse,id]) items
f (insertEv x item)
insertEv x [] = [[x]]
insertEv x l@(y:ys) = (x:l) : map (y:) (insertEv x ys)
 
foldlZipWith::(a -> b -> c) -> (d -> c -> d) -> d -> [a] -> [b] -> d
foldlZipWith _ _ u [] _ = u
foldlZipWith _ _ u _ [] = u
foldlZipWith f g u (x:xs) (y:ys) = foldlZipWith f g (g u (f x y)) xs ys
 
foldl1ZipWith::(a -> b -> c) -> (c -> c -> c) -> [a] -> [b] -> c
foldl1ZipWith _ _ [] _ = error "First list is empty"
foldl1ZipWith _ _ _ [] = error "Second list is empty"
foldl1ZipWith f g (x:xs) (y:ys) = foldlZipWith f g (f x y) xs ys
 
multAdd::(a -> b -> c) -> (c -> c -> c) -> [[a]] -> [[b]] -> [[c]]
multAdd f g xs ys = map (\us -> foldl1ZipWith (\u vs -> map (f u) vs) (zipWith g) us ys) xs
 
mult:: Num a => [[a]] -> [[a]] -> [[a]]
mult xs ys = multAdd (*) (+) xs ys
 
elemPos::[[a]] -> Int -> Int -> a
elemPos ms i j = (ms !! i) !! j
 
prod:: Num a => ([[a]] -> Int -> Int -> a) -> [[a]] -> [Int] -> a
prod f ms = product.zipWith (f ms) [0..]
 
s_determinant:: Num a => ([[a]] -> Int -> Int -> a) -> [[a]] -> [([Int],Int)] -> a
s_determinant f ms = sum.map (\(is,s) -> fromIntegral s * prod f ms is)
 
elemCramerPos::Int -> Int -> [[a]] -> [[a]] -> Int -> Int -> a
elemCramerPos l k ks ms i j = if j /= l then elemPos ms i j else elemPos ks i k
 
solveCramer:: [[Rational]] -> [[Rational]] -> [[Rational]]
solveCramer ms ks = xs
where
xs | d /= 0 = go (reverse [0..pred.length.head $ ks])
| otherwise = []
go (u:us) = foldl glue (col u) us
glue us u = zipWith (\ys (y:_) -> y:ys) us (col u)
col k = map (\l -> [(/d) $ s_determinant (elemCramerPos l k ks) ms ps]) $ ls
ls = [0..pred.length $ ms]
ps = s_permutations ls
d = s_determinant elemPos ms ps
matI::(Num a) => Int -> [[a]]
matI n = [ [fromIntegral.fromEnum $ i == j | i <- [1..n]] | j <- [1..n]]
 
task::[[Rational]] -> [[Rational]] -> IO()
task a b = do
let x = solveCramer a b
let u = map (map fromRational) x
let y = mult a x
let identity = matI (length x)
let a1 = solveCramer a identity
let h = mult a a1
let z = mult a1 b
putStrLn "a ="
mapM_ print a
putStrLn "b ="
mapM_ print b
putStrLn "solve: a * x = b => x = solveCramer a b ="
mapM_ print x
putStrLn "u = fromRationaltoDouble x ="
mapM_ print u
putStrLn "verification: y = a * x = mult a x ="
mapM_ print y
putStrLn $ "test: y == b = "
print $ y == b
putStrLn "identity matrix: identity ="
mapM_ print identity
putStrLn "find: a1 = inv(a) => solve: a * a1 = identity => a1 = solveCramer a identity ="
mapM_ print a1
putStrLn "verification: h = a * a1 = mult a a1 ="
mapM_ print h
putStrLn $ "test: h == identity = "
print $ h == identity
putStrLn "z = a1 * b = mult a1 b ="
mapM_ print z
putStrLn "test: z == x ="
print $ z == x
 
main = do
let a = [[2,-1, 5, 1]
,[3, 2, 2,-6]
,[1, 3, 3,-1]
,[5,-2,-3, 3]]
let b = [[-3], [-32], [-47], [49]]
task a b
 
Output:
a =
[2 % 1,(-1) % 1,5 % 1,1 % 1]
[3 % 1,2 % 1,2 % 1,(-6) % 1]
[1 % 1,3 % 1,3 % 1,(-1) % 1]
[5 % 1,(-2) % 1,(-3) % 1,3 % 1]
b =
[(-3) % 1]
[(-32) % 1]
[(-47) % 1]
[49 % 1]
solve: a * x = b => x = solveCramer a b =
[2 % 1]
[(-12) % 1]
[(-4) % 1]
[1 % 1]
u = fromRationaltoDouble x =
[2.0]
[-12.0]
[-4.0]
[1.0]
verification: y = a * x = mult a x =
[(-3) % 1]
[(-32) % 1]
[(-47) % 1]
[49 % 1]
test: y == b = 
True
identity matrix: identity =
[1 % 1,0 % 1,0 % 1,0 % 1]
[0 % 1,1 % 1,0 % 1,0 % 1]
[0 % 1,0 % 1,1 % 1,0 % 1]
[0 % 1,0 % 1,0 % 1,1 % 1]
find: a1 = inv(a) => solve: a * a1 = identity => a1 = solveCramer a identity =
[4 % 171,11 % 171,10 % 171,8 % 57]
[(-55) % 342,(-23) % 342,119 % 342,2 % 57]
[107 % 684,(-5) % 684,11 % 684,(-7) % 114]
[7 % 684,(-109) % 684,103 % 684,7 % 114]
verification: h = a * a1 = mult a a1 =
[1 % 1,0 % 1,0 % 1,0 % 1]
[0 % 1,1 % 1,0 % 1,0 % 1]
[0 % 1,0 % 1,1 % 1,0 % 1]
[0 % 1,0 % 1,0 % 1,1 % 1]
test: h == identity = 
True
z = a1 * b = mult a1 b =
[2 % 1]
[(-12) % 1]
[(-4) % 1]
[1 % 1]
test: z == x =
True

Version 3[edit]

 
import Data.List
 
determinant::(Fractional a, Ord a) => [[a]] -> a
determinant ls = if null ls then 0 else pivot 1 (zip ls [(0::Int)..])
where
good rs ts = (abs.head.fst $ ts) <= (abs.head.fst $ rs)
go us (vs,i) = if v == 0 then (ws,i) else (zipWith (\x y -> y - x*v) us ws,i)
where (v,ws) = (head $ vs,tail vs)
change i (ys:zs) = map (\xs -> if (==i).snd $ xs then ys else xs) zs
pivot d [] = d
pivot d zs@((_,j):ys) = if 0 == u then 0 else pivot e ws
where
e = if i == j then u*d else -u*d
((u:us),i) = foldl1 (\rs ts -> if good rs ts then rs else ts) zs
ws = map (go (map (/u) us)) $ if i == j then ys else change i zs
 
solveCramer::(Fractional a, Ord a) => [[a]] -> [[a]] -> [[a]]
solveCramer as bs = if 0 == d then [] else ans bs
where
d = determinant as
ans = transpose.map go.transpose
where
ms = zip [0..] (transpose as)
go us = [ (/d) $ determinant [if i /= j then vs else us | (j,vs) <- ms] | (i,_) <- ms]
 
matI::(Num a) => Int -> [[a]]
matI n = [ [fromIntegral.fromEnum $ i == j | i <- [1..n]] | j <- [1..n]]
 
foldlZipWith::(a -> b -> c) -> (d -> c -> d) -> d -> [a] -> [b] -> d
foldlZipWith _ _ u [] _ = u
foldlZipWith _ _ u _ [] = u
foldlZipWith f g u (x:xs) (y:ys) = foldlZipWith f g (g u (f x y)) xs ys
 
foldl1ZipWith::(a -> b -> c) -> (c -> c -> c) -> [a] -> [b] -> c
foldl1ZipWith _ _ [] _ = error "First list is empty"
foldl1ZipWith _ _ _ [] = error "Second list is empty"
foldl1ZipWith f g (x:xs) (y:ys) = foldlZipWith f g (f x y) xs ys
 
multAdd::(a -> b -> c) -> (c -> c -> c) -> [[a]] -> [[b]] -> [[c]]
multAdd f g xs ys = map (\us -> foldl1ZipWith (\u vs -> map (f u) vs) (zipWith g) us ys) xs
 
mult:: Num a => [[a]] -> [[a]] -> [[a]]
mult xs ys = multAdd (*) (+) xs ys
 
task::[[Rational]] -> [[Rational]] -> IO()
task a b = do
let x = solveCramer a b
let u = map (map fromRational) x
let y = mult a x
let identity = matI (length x)
let a1 = solveCramer a identity
let h = mult a a1
let z = mult a1 b
putStrLn "a ="
mapM_ print a
putStrLn "b ="
mapM_ print b
putStrLn "solve: a * x = b => x = solveCramer a b ="
mapM_ print x
putStrLn "u = fromRationaltoDouble x ="
mapM_ print u
putStrLn "verification: y = a * x = mult a x ="
mapM_ print y
putStrLn $ "test: y == b = "
print $ y == b
putStrLn "identity matrix: identity ="
mapM_ print identity
putStrLn "find: a1 = inv(a) => solve: a * a1 = identity => a1 = solveCramer a identity ="
mapM_ print a1
putStrLn "verification: h = a * a1 = mult a a1 ="
mapM_ print h
putStrLn $ "test: h == identity = "
print $ h == identity
putStrLn "z = a1 * b = mult a1 b ="
mapM_ print z
putStrLn "test: z == x ="
print $ z == x
 
main = do
let a = [[2,-1, 5, 1]
,[3, 2, 2,-6]
,[1, 3, 3,-1]
,[5,-2,-3, 3]]
let b = [[-3], [-32], [-47], [49]]
task a b
 
Output:
a =
[2 % 1,(-1) % 1,5 % 1,1 % 1]
[3 % 1,2 % 1,2 % 1,(-6) % 1]
[1 % 1,3 % 1,3 % 1,(-1) % 1]
[5 % 1,(-2) % 1,(-3) % 1,3 % 1]
b =
[(-3) % 1]
[(-32) % 1]
[(-47) % 1]
[49 % 1]
solve: a * x = b => x = solveCramer a b =
[2 % 1]
[(-12) % 1]
[(-4) % 1]
[1 % 1]
u = fromRationaltoDouble x =
[2.0]
[-12.0]
[-4.0]
[1.0]
verification: y = a * x = mult a x =
[(-3) % 1]
[(-32) % 1]
[(-47) % 1]
[49 % 1]
test: y == b = 
True
identity matrix: identity =
[1 % 1,0 % 1,0 % 1,0 % 1]
[0 % 1,1 % 1,0 % 1,0 % 1]
[0 % 1,0 % 1,1 % 1,0 % 1]
[0 % 1,0 % 1,0 % 1,1 % 1]
find: a1 = inv(a) => solve: a * a1 = identity => a1 = solveCramer a identity =
[4 % 171,11 % 171,10 % 171,8 % 57]
[(-55) % 342,(-23) % 342,119 % 342,2 % 57]
[107 % 684,(-5) % 684,11 % 684,(-7) % 114]
[7 % 684,(-109) % 684,103 % 684,7 % 114]
verification: h = a * a1 = mult a a1 =
[1 % 1,0 % 1,0 % 1,0 % 1]
[0 % 1,1 % 1,0 % 1,0 % 1]
[0 % 1,0 % 1,1 % 1,0 % 1]
[0 % 1,0 % 1,0 % 1,1 % 1]
test: h == identity = 
True
z = a1 * b = mult a1 b =
[2 % 1]
[(-12) % 1]
[(-4) % 1]
[1 % 1]
test: z == x =
True

J[edit]

Implementation:

cramer=:4 :0
A=. x [ b=. y
det=. -/ .*
A %~&det (i.#A) b"_`[`]}&.|:"0 2 A
)

Task data:

A=: _&".;._2]t=: 0 :0
2 -1 5 1
3 2 2 -6
1 3 3 -1
5 -2 -3 3
)
 
b=: _3 _32 _47 49

Task example:

   A cramer b
2 _12 _4 1

JavaScript[edit]

 
var matrix = [
[2, -1, 5, 1],
[3, 2, 2, -6],
[1, 3, 3, -1],
[5, -2, -3, 3]
];
var freeTerms = [-3, -32, -47, 49];
 
var result = cramersRule(matrix,freeTerms);
console.log(result);
 
/**
* Compute Cramer's Rule
* @param {array} matrix x,y,z, etc. terms
* @param {array} freeTerms
* @return {array} solution for x,y,z, etc.
*/

function cramersRule(matrix,freeTerms) {
var det = detr(matrix),
returnArray = [],
i,
tmpMatrix;
 
for(i=0; i < matrix[0].length; i++) {
var tmpMatrix = insertInTerms(matrix, freeTerms,i)
returnArray.push(detr(tmpMatrix)/det)
}
return returnArray;
}
 
/**
* Inserts single dimensional array into
* @param {array} matrix multidimensional array to have ins inserted into
* @param {array} ins single dimensional array to be inserted vertically into matrix
* @param {array} at zero based offset for ins to be inserted into matrix
* @return {array} New multidimensional array with ins replacing the at column in matrix
*/

function insertInTerms(matrix, ins, at) {
var tmpMatrix = clone(matrix),
i;
for(i=0; i < matrix.length; i++) {
tmpMatrix[i][at] = ins[i];
}
return tmpMatrix;
}
/**
* Compute the determinate of a matrix. No protection, assumes square matrix
* function borrowed, and adapted from MIT Licensed numericjs library (www.numericjs.com)
* @param {array} m Input Matrix (multidimensional array)
* @return {number} result rounded to 2 decimal
*/

function detr(m) {
var ret = 1,
k,
A=clone(m),
n=m[0].length,
alpha;
 
for(var j =0; j < n-1; j++) {
k=j;
for(i=j+1;i<n;i++) { if(Math.abs(A[i][j]) > Math.abs(A[k][j])) { k = i; } }
if(k !== j) {
temp = A[k]; A[k] = A[j]; A[j] = temp;
ret *= -1;
}
Aj = A[j];
for(i=j+1;i<n;i++) {
Ai = A[i];
alpha = Ai[j]/Aj[j];
for(k=j+1;k<n-1;k+=2) {
k1 = k+1;
Ai[k] -= Aj[k]*alpha;
Ai[k1] -= Aj[k1]*alpha;
}
if(k!==n) { Ai[k] -= Aj[k]*alpha; }
}
if(Aj[j] === 0) { return 0; }
ret *= Aj[j];
}
return Math.round(ret*A[j][j]*100)/100;
}
 
/**
* Clone two dimensional Array using ECMAScript 5 map function and EcmaScript 3 slice
* @param {array} m Input matrix (multidimensional array) to clone
* @return {array} New matrix copy
*/

function clone(m) {
return m.map(function(a){return a.slice();});
}
 
 
Output:
[ 2, -12, -4, 1 ]

Julia[edit]

function cramer_solve(A, v)
return [ begin
B = copy(A)
det((B[:, i] = v; B))
end for i = 1 : length(v) ] ./ det(A)
end
 
A = [
2 -1 5 1;
3 2 2 -6;
1 3 3 -1;
5 -2 -3 3;
]
v = [-3, -32, -47, 49]
 
map(println, round(cramer_solve(A, v), 14))

Note that it is entirely impractical to use Cramer's rule in this situation. It would be much better to use the built-in operator for solving linear systems. Assuming that the coefficient matrix and constant vector are defined as before, the solution vector is given by:

A \ v
Output:
2.0
-12.0
-4.0
1.0

Mathematica[edit]

crule[m_, b_] := Module[{d = Det[m], a},
Table[a = m; a[[All, k]] = b; Det[a]/d, {k, Length[m]}]]
 
crule[{
{2, -1, 5, 1},
{3, 2, 2, -6},
{1, 3, 3, -1},
{5, -2, -3, 3}
} , {-3, -32, -47, 49}]
Output:
{2,-12,-4,1}

PARI/GP[edit]

 
M = [2,-1,5,1;3,2,2,-6;1,3,3,-1;5,-2,-3,3];
V = Col([-3,-32,-47,49]);
 
matadjoint(M) * V / matdet(M)
 

Output:

[2, -12, -4, 1]~

Perl[edit]

use Math::Matrix;
 
sub cramers_rule {
my ($A, $terms) = @_;
my @solutions;
my $det = $A->determinant;
foreach my $i (0 .. $#{$A}) {
my $Ai = $A->clone;
foreach my $j (0 .. $#{$terms}) {
$Ai->[$j][$i] = $terms->[$j];
}
push @solutions, $Ai->determinant / $det;
}
@solutions;
}
 
my $matrix = Math::Matrix->new(
[2, -1, 5, 1],
[3, 2, 2, -6],
[1, 3, 3, -1],
[5, -2, -3, 3],
);
 
my $free_terms = [-3, -32, -47, 49];
my ($w, $x, $y, $z) = cramers_rule($matrix, $free_terms);
 
print "w = $w\n";
print "x = $x\n";
print "y = $y\n";
print "z = $z\n";
Output:
w = 2
x = -12
y = -4
z = 1

Perl 6[edit]

sub det(@matrix) {
my @a = @matrix.map: { [|$_] };
my $sign = +1;
my $pivot = 1;
for ^@a -> $k {
my @r = ($k+1 .. @a.end);
my $previous-pivot = $pivot;
if 0 == ($pivot = @a[$k][$k]) {
(my $s = @r.first: { @a[$_][$k] != 0 }) // return 0;
(@a[$s],@a[$k]) = (@a[$k], @a[$s]);
my $pivot = @a[$k][$k];
$sign = -$sign;
}
for @r X @r -> ($i, $j) {
((@a[$i][$j] *= $pivot) -= @a[$i][$k]*@a[$k][$j]) /= $previous-pivot;
}
}
$sign * $pivot
}
 
sub cramers_rule(@A, @terms) {
gather for ^@A -> $i {
my @Ai = @A.map: { [|$_] };
for ^@terms -> $j {
@Ai[$j][$i] = @terms[$j];
}
take det(@Ai);
} »/» det(@A);
}
 
my @matrix = (
[2, -1, 5, 1],
[3, 2, 2, -6],
[1, 3, 3, -1],
[5, -2, -3, 3],
);
 
my @free_terms = (-3, -32, -47, 49);
my ($w, $x, $y, $z) = |cramers_rule(@matrix, @free_terms);
 
say "w = $w";
say "x = $x";
say "y = $y";
say "z = $z";
Output:
w = 2
x = -12
y = -4
z = 1

Phix[edit]

Translation of: C

The copy-on-write semantics of Phix really shine here; because there is no explicit return/re-assign, you can treat parameters as a private workspace, confident in the knowledge that the updated version will be quietly discarded; all the copying and freeing of the C version is automatic/unnecessary here.

constant inf = 1e300*1e300,
nan = -(inf/inf)
 
function det(sequence a)
atom res = 1
 
for j=1 to length(a) do
integer i_max = j
for i=j+1 to length(a) do
if a[i][j] > a[i_max][j] then
i_max = i
end if
end for
if i_max != j then
{a[i_max],a[j]} = {a[j],a[i_max]}
res *= -1
end if
 
if abs(a[j][j]) < 1e-12 then
puts(1,"Singular matrix!")
return nan
end if
 
for i=j+1 to length(a) do
atom mult = -a[i][j] / a[j][j]
for k=1 to length(a) do
a[i][k] += mult * a[j][k]
end for
end for
end for
 
for i=1 to length(a) do
res *= a[i][i]
end for
return res
end function
 
function cramer_solve(sequence a, atom det_a, sequence b, integer var)
for i=1 to length(a) do
a[i][var] = b[i]
end for
return det(a)/det_a
end function
 
sequence a = {{2,-1, 5, 1},
{3, 2, 2,-6},
{1, 3, 3,-1},
{5,-2,-3, 3}},
b = {-3,-32,-47,49}
integer det_a = det(a)
for i=1 to length(a) do
printf(1, "%7.3f\n", cramer_solve(a, det_a, b, i))
end for
Output:
  2.000
-12.000
 -4.000
  1.000

Prolog[edit]

Works with: GNU Prolog
removeElement([_|Tail], 0, Tail).
removeElement([Head|Tail], J, [Head|X]) :-
J_2 is J - 1,
removeElement(Tail, J_2, X).
 
removeColumn([], _, []).
removeColumn([Matrix_head|Matrix_tail], J, [X|Y]) :-
removeElement(Matrix_head, J, X),
removeColumn(Matrix_tail, J, Y).
 
removeRow([_|Matrix_tail], 0, Matrix_tail).
removeRow([Matrix_head|Matrix_tail], I, [Matrix_head|X]) :-
I_2 is I - 1,
removeRow(Matrix_tail, I_2, X).
 
cofactor(Matrix, I, J, X) :-
removeRow(Matrix, I, Matrix_2),
removeColumn(Matrix_2, J, Matrix_3),
det(Matrix_3, Y),
X is (-1) ** (I + J) * Y.
 
det_summand(_, _, [], 0).
det_summand(Matrix, J, B, X) :-
B = [B_head|B_tail],
cofactor(Matrix, 0, J, Z),
J_2 is J + 1,
det_summand(Matrix, J_2, B_tail, Y),
X is B_head * Z + Y.
 
det([[X]], X).
det(Matrix, X) :-
Matrix = [Matrix_head|_],
det_summand(Matrix, 0, Matrix_head, X).
 
replaceElement([_|Tail], 0, New, [New|Tail]).
replaceElement([Head|Tail], J, New, [Head|Y]) :-
J_2 is J - 1,
replaceElement(Tail, J_2, New, Y).
 
replaceColumn([], _, _, []).
replaceColumn([Matrix_head|Matrix_tail], J, [Column_head|Column_tail], [X|Y]) :-
replaceElement(Matrix_head, J, Column_head, X),
replaceColumn(Matrix_tail, J, Column_tail, Y).
 
cramerElements(_, B, L, []) :- length(B, L).
cramerElements(A, B, J, [X_J|Others]) :-
replaceColumn(A, J, B, A_J),
det(A_J, Det_A_J),
det(A, Det_A),
X_J is Det_A_J / Det_A,
J_2 is J + 1,
cramerElements(A, B, J_2, Others).
 
cramer(A, B, X) :- cramerElements(A, B, 0, X).
 
results(X) :-
A = [
[2, -1, 5, 1],
[3, 2, 2, -6],
[1, 3, 3, -1],
[5, -2, -3, 3]
],
B = [-3, -32, -47, 49],
cramer(A, B, X).
Output:
| ?- results(X).
X = [2.0,-12.0,-4.0,1.0] ? 
yes

Python[edit]

 
# a simple implementation using numpy
from numpy import linalg
 
A=[[2,-1,5,1],[3,2,2,-6],[1,3,3,-1],[5,-2,-3,3]]
B=[-3,-32,-47,49]
C=[[2,-1,5,1],[3,2,2,-6],[1,3,3,-1],[5,-2,-3,3]]
X=[]
for i in range(0,len(B)):
for j in range(0,len(B)):
C[j][i]=B[j]
if i>0:
C[j][i-1]=A[j][i-1]
X.append(round(linalg.det(C)/linalg.det(A),1))
 
print('w=%s'%X[0],'x=%s'%X[1],'y=%s'%X[2],'z=%s'%X[3])
 
Output:
w=2.0 x=-12.0 y=-4.0 z=1.0

REXX[edit]

/* Use Cramer's rule to compute solutions of given linear equations  */
Numeric Digits 20
names='w x y z'
M=' 2 -1 5 1',
' 3 2 2 -6',
' 1 3 3 -1',
' 5 -2 -3 3'
v=' -3',
'-32',
'-47',
' 49'
Call mk_mat(m) /* M -> a.i.j */
Do j=1 To dim /* Show the input */
ol=''
Do i=1 To dim
ol=ol format(a.i.j,6)
End
ol=ol format(word(v,j),6)
Say ol
End
Say copies('-',35)
 
d=det(m) /* denominator determinant */
 
Do k=1 To dim /* construct nominator matrix */
Do j=1 To dim
Do i=1 To dim
If i=k Then
b.i.j=word(v,j)
Else
b.i.j=a.i.j
End
End
Call show_b
d.k=det(mk_str()) /* numerator determinant */
Say word(names,k) '=' d.k/d /* compute value of variable k */
End
Exit
 
mk_mat: Procedure Expose a. dim /* Turn list into matrix a.i.j */
Parse Arg list
dim=sqrt(words(list))
k=0
Do j=1 To dim
Do i=1 To dim
k=k+1
a.i.j=word(list,k)
End
End
Return
 
mk_str: Procedure Expose b. dim /* Turn matrix b.i.j into list */
str=''
Do j=1 To dim
Do i=1 To dim
str=str b.i.j
End
End
Return str
 
show_b: Procedure Expose b. dim /* show numerator matrix */
do j=1 To dim
ol=''
Do i=1 To dim
ol=ol format(b.i.j,6)
end
Call dbg ol
end
Return
 
det: Procedure /* compute determinant */
Parse Arg list
n=words(list)
call dbg 'det:' list
do dim=1 To 10
If dim**2=n Then Leave
End
call dbg 'dim='dim
If dim=2 Then Do
det=word(list,1)*word(list,4)-word(list,2)*word(list,3)
call dbg 'det=>'det
Return det
End
k=0
Do j=1 To dim
Do i=1 To dim
k=k+1
a.i.j=word(list,k)
End
End
Do j=1 To dim
ol=j
Do i=1 To dim
ol=ol format(a.i.j,6)
End
call dbg ol
End
det=0
Do i=1 To dim
ol=''
Do j=2 To dim
Do ii=1 To dim
If ii<>i Then
ol=ol a.ii.j
End
End
call dbg 'i='i 'ol='ol
If i//2 Then
det=det+a.i.1*det(ol)
Else
det=det-a.i.1*det(ol)
End
Call dbg 'det=>>>'det
Return det
 
dbg: Return
Output:
      2     -1      5      1     -3
      3      2      2     -6    -32
      1      3      3     -1    -47
      5     -2     -3      3     49
-----------------------------------
w = 2
x = -12
y = -4
z = 1

Ruby[edit]

require 'matrix'
 
def cramers_rule(a, terms)
raise ArgumentError, " Matrix not square" unless a.square?
cols = a.to_a.transpose
cols.each_index.map do |i|
c = cols.dup
c[i] = terms
Matrix.columns(c).det / a.det
end
end
 
matrix = Matrix[
[2, -1, 5, 1],
[3, 2, 2, -6],
[1, 3, 3, -1],
[5, -2, -3, 3],
]
 
vector = [-3, -32, -47, 49]
puts cramers_rule(matrix, vector)
Output:
2
-12
-4
1

Sidef[edit]

func det(a) {
a = a.map{.map{_}}
var sign = +1
var pivot = 1
 
for k in ^a {
var r = (k+1 .. a.end)
var previous_pivot = pivot
 
if ((pivot = a[k][k]) == 0) {
a.swap(r.first_by { a[_][k] != 0 } \\ return 0, k)
pivot = a[k][k]
sign.neg!
}
 
for i,j in (r ~X r) {
a[i][j] *= pivot ->
-= a[i][k]*a[k][j] ->
/= previous_pivot
}
}
sign * pivot
}
 
func cramers_rule(A, terms) {
gather {
for i in ^A {
var Ai = A.map{.map{_}}
for j in ^terms {
Ai[j][i] = terms[j]
}
take(det(Ai))
}
} »/» det(A)
}
 
var matrix = [
[2, -1, 5, 1],
[3, 2, 2, -6],
[1, 3, 3, -1],
[5, -2, -3, 3],
]
 
var free_terms = [-3, -32, -47, 49]
var (w, x, y, z) = cramers_rule(matrix, free_terms)...
 
say "w = #{w}"
say "x = #{x}"
say "y = #{y}"
say "z = #{z}"
Output:
w = 2
x = -12
y = -4
z = 1

zkl[edit]

Using the GNU Scientific Library, we define the values:

var [const] GSL=Import("zklGSL");	// libGSL (GNU Scientific Library)
A:=GSL.Matrix(4,4).set(2,-1, 5, 1,
3, 2, 2,-6,
1, 3, 3,-1,
5,-2,-3, 3);
b:=GSL.Vector(4).set(-3,-32,-47,49);

First, just let GSL solve:

A.AxEQb(b).format().println();

Actually implement Cramer's rule:

Translation of: Julia
fcn cramersRule(A,b){
b.len().pump(GSL.Vector(b.len()),'wrap(i){ // put calculations into new Vector
A.copy().setColumn(i,b).det();
}).close()/A.det();
}
cramersRule(A,b).format().println();
Output:
2.00,-12.00,-4.00,1.00