* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Lecture Notes 1
Location arithmetic wikipedia , lookup
List of important publications in mathematics wikipedia , lookup
Proofs of Fermat's little theorem wikipedia , lookup
Strähle construction wikipedia , lookup
Elementary mathematics wikipedia , lookup
Mathematics of radio engineering wikipedia , lookup
System of polynomial equations wikipedia , lookup
ANALYTICAL MATHEMATICS FOR APPLICATIONS 2017 LECTURE NOTES 1 ISSUED 10 FEBRUARY 2017 1. Two linear equations in two unknowns: Cramer’s rule We want to solve a system of linear equations a1 x + b1 y = c1 a2 x + b2 y = c2 where a1 , a2 , a3 , a4 are either integer or rational or real numbers. Consider a matrix (square table of numbers) corresponding to this system, ( ) a1 b1 A= . a2 b2 Definition 1.1. The number det A = a1 b2 − a2 b1 is called the determinant of the matrix A. Consider also the following matrices ( ) c1 b1 Ax = , c2 b2 ( Ay = a1 a2 c1 c2 ) . Theorem 1.2 (Cramer’s rule). Assume that the determinant det A ̸= 0. Then a solution of the system of equations is given by the following formulas: x= det Ax , det A y= det Ay . det A Proof. In the system of equations, replace x by det Ax c 1 b2 − c 2 b1 = , det A a 1 b2 − a 2 b1 and y by det Ay a1 c2 − a2 c1 = . det A a1 b2 − a2 b1 Then a direct computation shows that the left-hand sides of both equations evaluate to the right-hand sides. Corollary 1.3. If coefficients of the system of equations are rational numbers and det A ̸= 0, then there is a solution (x, y) of the system such that x and y are rational numbers. 1 2 ISSUED 10 FEBRUARY 2017 The case when the determinant det A equals to 0 is easy to analyze and conclude that then the system of equations either does not have any solutions or has infinitely many solutions which coincide with the set of all solutions of any of the two equations. Cramer’s rule generalizes to any system of linear equations in which the number of equations equals the number of unknowns. For this purpose we need to generalize the concept of a determinant. We will do this in a later section. The example we just considered shows the usefulness of using matrices in the questions about solving linear systems. This prompts a development of a theory of matrices. 2. Matrices An (m × n)-matrix A is a rectangular table of rational or numbers having m rows and n columns: a11 a12 · · · a1n a21 a22 · · · a2n . A= ··· ··· ··· ··· am1 am2 · · · amn The numbers aij , where i = 1, . . . m, j = 1, . . . n, are called elements of A. We will often write a matrix indicating just its “general” element, A = ∥aij ∥1≤i≤m, 1≤j≤n , or simply A = ∥aij ∥. 3. Determinants Let A = ∥aij ∥1≤i,j≤n be a square (n × n)-matrix. Denote by Aij the matrix obtained from A by striking out of A the row i and the column j. Definition 3.1. The determinant det A of A is a number defined by induction as follows. If n = 1 then the matrix A contains just one element a11 and we set det A = a11 . Assume that n > 1 and that we defined determinants of (n − 1) × (n − 1)-matrices. Fix any number i such that 1 ≤ i ≤ n. Then define det A = (−1)i+1 ai1 det Ai1 + · · · + (−1)i+j aij det Aij + · · · + (−1)i+n ain det Ain . It can be proved that det A does not depend on the choice of the number i, i.e., for any 1 ≤ i ≤ n the number det A will be the same. Example 3.2. Let ( A= a11 a21 a12 a22 ) . We already know (see Lecture Notes 1) that det A = a11 a22 −a21 a22 . Let us confirm that using Definition 3.1, in which we take i = 1: det A = (−1)1+1 a11 det A11 + (−1)1+2 a12 det A12 = a11 a22 − a12 a21 . Example 3.3. Let us consider a numerical 1 2 A= 2 1 3 2 example of a (3 × 3)-matrix 3 3 . 1 LECTURE NOTES 1 Choose again in Definition 3.1 the row i = 1. ( ) ( 1 3 2 det A = (−1)1+1 det + (−1)1+2 2 2 1 3 3 3 1 ) ( 2 3 1 2 = −5 + 14 + 3 = 12. Now choose the row i = 2. ( ) ( ) ( 2 3 1 3 1 det A = (−1)2+1 2 det + (−1)2+2 + (−1)2+3 3 2 1 3 1 3 2 2 + (−1)1+3 3 ) = ) = = 8 − 8 + 12 = 12. Exercise. Choose i = 3 and make sure that in this case the formula from Definition 3.1 again gives det A = 12. Let us now give another, non-inductive, definition of the determinant. Definition 3.4. Let i1 , i2 , . . . , in be a permutation of numbers 1, 2, . . . , n. An inversion in i1 , i2 , . . . , in is a pair (ik , il ) such that k < l and ik > il . For example, in the permutation 2, 3, 1 of 1, 2, 3 there are two inversions, (2, 1) and (3, 1). In an (n × n)-matrix A choose n different elements positioned in different rows and different columns, i.e., for any two elements aij , akℓ among these n, we have i ̸= k and j ̸= ℓ. Notice that n is the maximal number of elements with such property, call the set of these elements a configuration. One can easily check that there is exactly n! = 1 · 2 · · · n different configurations. Definition 3.5. The determinant det A of A is the sum over all configurations {ai1 j1 , ai2 j2 , . . . , ain jn } ∑ det A = (−1)σ(i1 ,...,in ;j1 ,...,jn ) ai1 j1 ai2 j2 · · · ain jn {ai1 j1 ,ai2 j2 ,...,ain jn } where σ(i1 , . . . , in ; j1 , . . . , jn ) is the number defined as follows. Consider the tworow matrix ( ) i1 i2 · · · in , j1 j2 · · · jn and re-order its columns so that the first row i1 , i2 . . . , in turns into 1, 2, . . . , n: ( ) 1 2 ··· n . r1 r2 · · · rn Then σ(i1 , . . . , in ; j1 , . . . , jn ) is defined as the number of inversions (Definition 3.4) in the permutation r1 , r2 , . . . , rn . Example 3.6. If ( ) a11 a12 , a21 a22 then there are exactly two different configurations, {a11 , a22 } and {a12 , a21 }. Hence, A= det A = (−1)σ(1,2;1,2) a11 a22 + (−1)σ(1,2;2,1) a12 a21 , where the numbers of inversions σ(1, 2; 1, 2) = 0 and σ(1, 2; 2, 1) = 1. We recover our original definition of the determinant of a (2 × 2)-matrix. If a11 a12 a13 A = a21 a22 a23 , a31 a32 a33 4 ISSUED 10 FEBRUARY 2017 then there are six different configurations, and the determinant det A is equal to a11 a22 a33 + a12 a23 a31 + a13 a21 a32 − a13 a22 a31 − a12 a21 a33 − a11 a23 a32 . 4. Cramer’s rule Now we are able to formulate the Cramer’s rule for general (n × n)-matrices. Consider the system of linear equations a11 X1 + a12 X2 + · · · + a1n Xn = b1 ··· ··· ··· (∗) an1 X1 + an2 X2 + · · · + ann Xn = bn As discussed in Lecture Notes 2, it can be represented in a form AX = b, where A = ∥aij ∥, X = (X1 , . . . Xn )T and b = (b1 , . . . bn )T . For each i = 1, 2, . . . , n denote by Axj the matrix obtained from A by replacing the column j, i.e., (a1j , a2j , . . . , anj )T by b = (b1 , . . . bn )T . Theorem 4.1 (Cramer’s rule). Solutions of the system (∗) are given by the following formulas det Axj for all j = 1, 2, . . . , n. xj = det A