Let Mn be the n×n matrix with entries as follows: for 1≤i≤n, mi,i=10; for 1≤i≤n−1, mi+1,i=mi,i+1=3; all other entries in Mn are zero. Let Dn be the determinant of matrix Mn. Then ∑n=1∞8Dn+11 can be represented as qp, where p and q are relatively prime positive integers. Find p+q.
Note: The determinant of the 1×1 matrix [a] is a, and the determinant of the 2×2 matrix [acbd]=ad−bc; for n≥2, the determinant of an n×n matrix with first row or first column a1a2a3…an is equal to a1C1−a2C2+a3C3−⋯+(−1)n+1anCn, where Ci is the determinant of the (n−1)×(n−1) matrix formed by eliminating the row and column containing ai.
解析
Solution
D1=10=10,D2=103310=(10)(10)−(3)(3)=91,D3=103031030310.
Using the expansionary/recursive definition of determinants (also stated in the problem):
This pattern repeats because the first element in the first row of Mn is always 10, the second element is always 3, and the rest are always 0. The ten element directly expands to 10Dn−1. The three element expands to 3 times the determinant of the the matrix formed from omitting the second column and first row from the original matrix. Call this matrix Xn. Xn has a first column entirely of zeros except for the first element, which is a three. A property of matrices is that the determinant can be expanded over the rows instead of the columns (still using the recursive definition as given in the problem), and the determinant found will still be the same. Thus, expanding over this first column yields 3Dn−2+0(other things)=3Dn−2. Thus, the 3det(Xn) expression turns into 9Dn−2. Thus, the equation Dn=10Dn−1−9Dn−2 holds for all n > 2.
This equation can be rewritten as Dn=10(Dn−1−Dn−2)+Dn−2. This version of the equation involves the difference of successive terms of a recursive sequence. Calculating D0 backwards from the recursive formula and D4 from the formula yields D0=1,D4=7381. Examining the differences between successive terms, a pattern emerges. D0=1=90, D1−D0=10−1=9=91, D2−D1=91−10=81=92, D3−D2=820−91=729=93, D4−D3=7381−820=6561=94. Thus, Dn=D0+91+92+...+9n=∑i=0n9i=9−1(1)(9n+1−1)=89n+1−1.
Thus, the desired sum is ∑n=1∞889n+1−1+11=∑n=1∞9n+1−1+11=∑n=1∞9n+11
This is an infinite geometric series with first term 811 and common ratio 91. Thus, the sum is 1−91811=98811=(81)(8)9=(9)(8)1=721.
Thus, p+q=1+72=073.
Solution 2 (Engineer's Induction - not rigorous)
Solution 3 (Uses first half of Solution 1)
From Solution 1, Dn=10Dn−1−9Dn−2. From Solution 1, we also know D2=91. We can manually compute D1=10. Iterating backwards, D0=1.
The characteristic polynomial of this recurrence is
x2−10x+9,
which has roots 1 and 9. Therefore, the explicit formula for Dn is:
Dn=α1+9nα2.
We can solve for α1 and α2 by setting n=0 and n=1 respectively. This gives the following equations:
α1+α2=1,α1+9α2=10.
Solving these equations, we find
α1=−81,α2=89.
Substituting these back into the explicit formula, we find
Dn=89n+1−1.
Notice that
8Dn+1+11=9−(n+1).
The summation asks us to find the sum of a geometric series with ratio 91 and first term 811. The sum of this series is
1−91811=98811=721.
Thus, the answer is 73
Faster Alternate
Optionally, one can notice that the recursive relation is a first order non-homogenous relation in the form an=ran−1+c. So, to solve this (and all other first orders), we first need to find a point L for which
L=rL+c
in which we have
L=9L+1L=−81
Such L is called the Equilibrium Point of a recurrence relation. Its property satisfies:
Dn−L=r(Dn−1−L)Dn+81=9(Dn−1+81
Let Vn=Dn+81. The result is a geometric sequence with first term 881. The explicit rule for bn is just bn=881⋅9n−1. Conversion back to Dn can be done quite easily.
Dn+81=89n+1Dn=89n+1−1
.
The explicit rule has been reached and one can simply continue as follows.