Download methods of solving a linear system – echelon form

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Maxwell's equations wikipedia , lookup

Unification (computer science) wikipedia , lookup

Two-body problem in general relativity wikipedia , lookup

Equations of motion wikipedia , lookup

BKL singularity wikipedia , lookup

Perturbation theory wikipedia , lookup

Calculus of variations wikipedia , lookup

Schwarzschild geodesics wikipedia , lookup

Differential equation wikipedia , lookup

Computational electromagnetics wikipedia , lookup

Exact solutions in general relativity wikipedia , lookup

Partial differential equation wikipedia , lookup

Transcript
1
โ— METHODS OF SOLVING A LINEAR SYSTEM โ€“ ECHELON FORM
๐‘ฅ + 3๐‘ฆ โˆ’ 2๐‘ง = 5
3๐‘ฅ + 5๐‘ฆ + 6๐‘ง = 7
2๐‘ฅ + 4๐‘ฆ + 3๐‘ง = 8
Consider the following linear system :
There are several algorithms for solving a system of linear equations.
1. Elimination of variables โ€“ no matrices
Substitution: The simplest method for solving a system of linear equations is to repeatedly eliminate variables.
2. Row reduction (Gaussian elimination) โ€“ Augmented matrix
In row reduction, the linear system is represented as an augmented matrix:
1
[3
2
3
5
4
โˆ’2 5
6 | 7]
3 8
This matrix is then modified using elementary row operations until it reaches reduced echelon form.
There are three types of elementary row operations:
Type 1: Interchange equations โ†’ Swap the positions of two rows.
Type 2: Multiply equation โ†’ Multiply a row by a nonzero scalar.
Type 3: Any of these equations can be replaced by a multiple of itself ± a multiple of another equation
without changing the outcome โ†’ Add to one row a scalar multiple of another.
The augmented matrix produced always represents a linear system that is equivalent to the original.
The following computation shows Gauss-Jordan elimination applied to the matrix above:
Combine two at the time to eliminate one variable (for example ๐‘ฅ from two equations) and then combine these
two to eliminate the second variable from one of them (for example ๐‘ฆ)
๐‘ฅ + 3๐‘ฆ โˆ’ 2๐‘ง = 5
3๐‘ฅ + 5๐‘ฆ + 6๐‘ง = 7
2๐‘ฅ + 4๐‘ฆ + 3๐‘ง = 8
1
[3
2
[๐‘…2 โˆ’ 3๐‘…1 ] ~
3 โˆ’2 5
5 6 | 7]
4 3 8
[๐‘…2 โˆ’ 2๐‘…3 ] ~
~
๐‘ฅ + 3๐‘ฆ โˆ’ 2๐‘ง = 5
โˆ’2๐‘ง = โˆ’4
โˆ’2๐‘ฆ + 7๐‘ง = โˆ’2
1
[0
0
3
0
โˆ’2
๐‘ฅ + 3๐‘ฆ โˆ’ 2๐‘ง = 5
โˆ’4๐‘ฆ + 12๐‘ง = โˆ’8
2๐‘ฅ + 4๐‘ฆ + 3๐‘ง = 8
1
[0
2
3
โˆ’4
4
[๐‘…3 โˆ’ 2๐‘…1 ]
โˆ’2 5
12 | โˆ’8]
3 8
๐‘…2 = ๐‘…3
๐‘…3 = โˆ’๐‘…2
โˆ’2 5
โˆ’2| โˆ’4]
7 โˆ’2
~
~
~
~
๐‘ฅ + 3๐‘ฆ โˆ’ 2๐‘ง = 5
โˆ’4๐‘ฆ + 12๐‘ง = โˆ’8
โˆ’2๐‘ฆ + 7๐‘ง = โˆ’2
1
[0
0
3
โˆ’4
โˆ’2
โˆ’2 5
12 | โˆ’8]
7 โˆ’2
๐‘ฅ + 3๐‘ฆ โˆ’ 2๐‘ง = 5
โˆ’2๐‘ฆ + 7๐‘ง = โˆ’2
2๐‘ง = 4
1
[0
0
3
โˆ’2
0
โˆ’2 5
7 | โˆ’2]
2 4
The last matrix is in reduced row echelon form โ€“ augmented matrix with the zeroes in the bottom left corner.
It represents the solution of the system of three linear equations.
๐’›=๐Ÿ
๐ฒ=๐Ÿ–
๐’™ = โˆ’๐Ÿ๐Ÿ“
(โˆ’2y + 2 โˆ™ 7 = โˆ’2 )
(๐‘ฅ + 3 โˆ™ 8 โˆ’ 2 โˆ™ 2 = 5)
2
โ— THE MEANING OF THE SOLUTIONS
THE A LINEAR SYSTEM MAY BEHAVE IN ANY OF THREE POSSIBLE WAYS
1. The system has a single unique solution.
2. The system has no solution.
3. The system has infinitely many solutions.
Geometric interpretation
2โ€“D
โˆŽ For a system involving two variables (x and y), each linear equation determines a line on the xy-plane. Because a
solution to a linear system must satisfy all of the equations, the solution set is the intersection of these lines, and is
hence either
unique solution: a single point
no solution: the empty set
infinitely many solutions: a line
parallel lines
The slopes are equal โ€“ no common points
coincident lines
The slopes are equal โ€“ one common point
โ‡’ all points are common
3โ€“D
โˆŽ For three variables, each linear equation determines a plane in 3-D space, and the solution set is the intersection
of these planes. Thus the solution set may be a plane, a line, a single point, or the empty set.
โˆŽ For n variables, each linear equation determines a hyperplane in n-D space. The solution set is the intersection
of these hyperplanes, which may be a flat of any dimension.
A ๐’ × ๐’ homogeneous system of linear equations has a unique solution (the trivial solution) if and only if its
determinant is non-zero. If this determinant is zero, then the system has an infinite number of solutions.
3
โ— SOLUTIONS OF A LINEAR SYSTEM USING AUGMENTED MATRIX IN ECHELON FORM
using row reduction (Gaussian elimination) for augmented matrix to get echelon form
System
a11x1 + a12x2 + a13x3 = b1
a 21x1 + a 22x2 + a23x3 = b2
a 31x1 + a32x2 + a33x3 = b3
Augmented form of the matrix of coefficients of the system
๐‘Ž11
๐‘Ž
[ 21
๐‘Ž31
๐‘Ž12
๐‘Ž22
๐‘Ž32
๐‘Ž13 ๐‘1
๐‘Ž23 | ๐‘2 ]
๐‘Ž33 ๐‘3
Echelon form
๐‘Ž
[0
0
๐‘ ๐‘ ๐‘‘
๐‘˜
โ†’ ๐‘ฆ&๐‘ฅ
๐‘’ ๐‘“ | ๐‘”] โ†’ ๐‘ง =
โ„Ž
0 โ„Ž ๐‘˜
๐’–๐’๐’Š๐’’๐’–๐’† ๐’”๐’๐’๐’–๐’•๐’Š๐’๐’:
๐’๐’ ๐’”๐’๐’๐’–๐’•๐’Š๐’๐’:
๐‘˜
โ„Ž โ‰ 0 โ†’ ๐‘ง =โ„Ž โ†’ ๐‘ฆ&๐‘ฅ
(๐‘˜ ๐‘š๐‘Ž๐‘ฆ ๐‘œ๐‘Ÿ ๐‘š๐‘Ž๐‘ฆ ๐‘›๐‘œ๐‘ก ๐‘๐‘’ 0)
โ„Ž = 0 ๐‘Ž๐‘›๐‘‘ ๐‘˜ โ‰  0 โ†’ 0 โˆ™ ๐‘ง = ๐‘˜ โ‰  0 ๐‘คโ„Ž๐‘–๐‘โ„Ž ๐‘–๐‘  ๐‘Ž๐‘๐‘ ๐‘ข๐‘Ÿ๐‘‘ โ†’ ๐‘กโ„Ž๐‘’๐‘Ÿ๐‘’ ๐‘–๐‘  ๐‘›๐‘œ ๐‘ ๐‘œ๐‘™๐‘ข๐‘ก๐‘–๐‘œ๐‘›
๐‘ ๐‘ฆ๐‘ ๐‘ก๐‘’๐‘š ๐‘–๐‘  ๐’Š๐’๐’„๐’๐’๐’”๐’Š๐’”๐’•๐’†๐’๐’•
๐’Š๐’๐’‡๐’Š๐’๐’Š๐’•๐’†๐’๐’š ๐’Ž๐’‚๐’๐’š ๐’”๐’๐’๐’–๐’•๐’Š๐’๐’๐’”: โ„Ž = 0 ๐‘Ž๐‘›๐‘‘ ๐‘˜ = 0 โ†’ ๐‘ง ๐‘๐‘Ž๐‘› ๐‘๐‘’ ๐‘Ž๐‘›๐‘ฆ ๐‘›๐‘ข๐‘š๐‘๐‘’๐‘Ÿ, ๐‘ ๐‘œ ๐‘ค๐‘’ ๐‘ค๐‘Ÿ๐‘–๐‘ก๐‘’: ๐‘ง = ๐‘ก ๐‘คโ„Ž๐‘’๐‘Ÿ๐‘’ ๐‘ก โˆˆ ๐‘…
๐‘ฅ = ๐‘ฅ(๐‘ก)
๐‘ฆ = ๐‘ฆ(๐‘ก)
๐‘ง=๐‘ก
Parametric representation of infinitely many solutions is not unique. We expressed the variable
which was not free (z) in terms of the parameter. We eliminated from row 3 variables x and y. It
does not have to be that way. We could eliminate z and y to get x, and then express y and z.
One can solve for any of the variables. Of course, the solution set will look different. However, it
will still represent the same solutions.
๐‘ข๐‘›๐‘–๐‘ž๐‘ข๐‘’ ๐‘ ๐‘œ๐‘™๐‘ข๐‘ก๐‘–๐‘œ๐‘›:
1 1
[0 2
0 0
๐‘›๐‘œ ๐‘ ๐‘œ๐‘™๐‘ข๐‘ก๐‘–๐‘œ๐‘›:
1
[0
0
๐‘–๐‘›๐‘“๐‘–๐‘›๐‘’๐‘ก๐‘’๐‘™๐‘ฆ ๐‘š๐‘Ž๐‘›๐‘ฆ ๐‘ ๐‘œ๐‘™๐‘ข๐‘ก๐‘–๐‘œ๐‘›๐‘ :
1 1
[0 2
0 0
1 1
2| 2] โ†’ ๐‘ง = 2
3 6
1 1 1
2 2| 2] โ†’ ๐‘ง โˆ™ 0 = 3
0 0 3
๐‘ฆ = โˆ’1 ๐‘ฅ = 0 ๐‘ƒ(0, โˆ’1,2)
๐‘Ž๐‘๐‘ ๐‘ข๐‘Ÿ๐‘‘
1 1
๐‘งโˆ™0=0 โ‡’ ๐‘ง =๐‘ก๐œ–๐‘…
2| 2] โ†’ โŸ
๐‘ก๐‘Ÿ๐‘ข๐‘’
๐‘“๐‘œ๐‘Ÿ ๐‘Ž๐‘›๐‘ฆ ๐‘ง
0 0
๐‘ฆ =1โˆ’๐‘ก
๐‘ฅ=0
4
โ— INTERSECTION OF TWO or MORE PLANES