CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Wiki > LU decomposition

LU decomposition

From CFD-Wiki

(Difference between revisions)
Jump to: navigation, search
(Important Considerations)
(Important Considerations)
Line 57: Line 57:
Here a Fortran code fragment for LU decomposition written by  D. Partenov
Here a Fortran code fragment for LU decomposition written by  D. Partenov
-
do j = 1, n
+
do j = 1, n
   do i = 1, j
   do i = 1, j
     if (i == 1) then
     if (i == 1) then
       a(i,j) = a(i, j)
       a(i,j) = a(i, j)
     else
     else
-
      suma = 0.0
+
      suma = 0.0
       do k = 1, i-1
       do k = 1, i-1
-
        suma = suma + a(i,k)*a(k,j)
+
        suma = suma + a(i,k)*a(k,j)
       end do
       end do
       a(i,j) = a(i,j) - suma
       a(i,j) = a(i,j) - suma
Line 83: Line 83:
     end do
     end do
   end if
   end if
-
end do
+
end do
-
 
+
y(1) = b(1)
-
y(1) = b(1)
+
do i = 2, n
-
 
+
-
do i = 2, n
+
   suma = 0.0
   suma = 0.0
   do j = 1, i-1
   do j = 1, i-1
Line 93: Line 91:
   end do
   end do
   y(i) = b(i) - suma
   y(i) = b(i) - suma
-
end do
+
end do
-
 
+
i = n
-
i = n
+
j = n
-
j = n
+
x(i) = y(i)/a(i,j)
-
x(i) = y(i)/a(i,j)
+
do s = 1, n -1
-
 
+
-
do s = 1, n -1
+
   i = n - s
   i = n - s
   suma = 0.0
   suma = 0.0
Line 106: Line 102:
   end do
   end do
   x(i) = (y(i) - suma)/a(i,i)
   x(i) = (y(i) - suma)/a(i,i)
-
end do
+
end do
== References ==
== References ==
*{{reference-book|author=Golub and Van Loan|year=1996|title=Matrix Computations|rest= 3rd edition, The Johns Hopkins University Press, Baltimore}}
*{{reference-book|author=Golub and Van Loan|year=1996|title=Matrix Computations|rest= 3rd edition, The Johns Hopkins University Press, Baltimore}}
*[http://en.wikipedia.org/wiki/LU_decomposition Wikipedia article ''LU decomposition]
*[http://en.wikipedia.org/wiki/LU_decomposition Wikipedia article ''LU decomposition]

Revision as of 18:38, 28 July 2007

Contents

Description

Consider the system of equations A\vec{x}=\vec{b}, where A is an n\times n nonsingular matrix. A may be decomposed into an lower triangular part L and an upper triangular part U that will lead us to a direct procedure for the solution of the original system. This decomposition procedure is quite useful when more than one right-hand side (more than one \vec{b}) is to be used.

The algorithm is relatively straightforward - first, we determine the upper and lower triangular parts:

A = LU.

Then,

 A \vec{x} = (LU) \vec{x} = L(U\vec{x})=L\vec{y},

where y=Ux. Once we solve the system

L\vec{y} = \vec{b},

we will be able to find the solution to the original system by solving

U\vec{x} = \vec{b}

The first solution is a foward substitution, while the second solution is a backward substitution. Both can be done efficiently once the factorization is available. The forward substitution may be expressed as


y_i  = {1 \over {l_{ii} }}\left( {b_i  - \sum\limits_{j = 1}^{i-1} {l_{ij} y_j } } \right),
i = 1,2,\ldots,n

and the backward substitution process may be expressed as


x_i  = {1 \over {u_{ii} }}\left( {y_i  - \sum\limits_{j = i + 1}^n {u_{ij} x_j } } \right),
i = n,n-1,\ldots,1.


LU decomposition essentially stores the operations of Gaussian elimination in "higher-level" form (see Golub and Van Loan), so repeated solutions using the same left-hand side are computed without repetition of operations that are independent of the right-hand side.

Algorithm

Add factorization here
Forward substitution

for k:=1 step until n do
for i:=1 step until k-1
b_k=b_k-l_{ki}b_{i}
end loop (i)
b_{k}=b_{k}/l_{kk}
end loop (k)

Backward substitution

for k:=n stepdown until 1 do
for i:=k+1 step until n
b_k=b_k-u_{ki}b_{i}
end loop (i)
x_{k}=b_{k}/u_{kk}
end loop (k)

Important Considerations

As with Gaussian elimination, LU decomposition is probably best used for relatively small, relatively non-sparse systems of equations (with small and non-sparse open to some interpretation). For larger and/or sparse problems, it would probably be best to either use an iterative method or use a direct solver package (e.g. DSCPACK) as opposed to writing one of your own.

If one has a single lefthand-side matrix and many right-hand side vectors, then LU decomposition would be a good solution procedure to consider. At the very least, it should be faster than solving each system separately with Gaussian elimination.

Here a Fortran code fragment for LU decomposition written by D. Partenov

do j = 1, n
 do i = 1, j
   if (i == 1) then
     a(i,j) = a(i, j)
   else
      suma = 0.0
     do k = 1, i-1
        suma = suma + a(i,k)*a(k,j)
     end do
     a(i,j) = a(i,j) - suma
   end if
 end do
 if ( j < n) then
   do s = 1, n-j
     i = j + s
     if (j == 1) then
       a(i,j) = a(i,j)/a(j,j)
     else
       suma = 0.0
       do k = 1, j-1
         suma = suma + a(i,k)*a(k,j)
       end do
       a(i,j) = (a(i,j) - suma)/a(j,j)
     end if
   end do
 end if
end do
y(1) = b(1)
do i = 2, n
 suma = 0.0
 do j = 1, i-1
   suma = suma + a(i,j)*y(j)
 end do
 y(i) = b(i) - suma
end do
i = n
j = n
x(i) = y(i)/a(i,j)
do s = 1, n -1
 i = n - s
 suma = 0.0
 do j = i+1, n
   suma = suma + a(i,j)*x(j)
 end do
 x(i) = (y(i) - suma)/a(i,i)
end do

References

My wiki