0

where $3\lt x_i\le 29, i=1,2$ and $10\le x_j\le 40, j=3,4.$

my work:

$x_1+x_2+...+x_6=110$ where $0\lt x_i-3\le 26, i=1,2$ and $1\le x_j-9\le 31, j=3,4$ and $0\lt x_6$

The number of solutions of the Eq. is the same as the number of the nonnegative integer solution of

$y_1+y_2+...+y_6=110-[3+3+9+9+1]=85$ where $0\lt y_i=x_i-3\le 26, i=1,2$ and $0\lt y_j=x_j-9\le 31, j=3,4$ and $0\lt y_6$

now I know that I have to get the total number of solutions and exclude the cases where $y_1,y_2\gt 26$ and $y_3,y_4\gt31 $

and this is the problem. I don't know how to do it. it's not like I'll keep assuming every case, so is there an effective way to do this.

StubbornAtom
  • 17,052
Leavei
  • 69
  • 1
    The answer is easy: infinite. Unless you add some restriction for $x_5$ – leonbloy May 14 '20 at 00:04
  • @leonbloy it is heavily implied that the OP is looking only for non-negative integer solutions. – JMoravitz May 14 '20 at 00:05
  • 1
    @ Leavei you have a very good start here and you should be very close to finishing. All that remains is to apply inclusion-exclusion. It does look like it will be a bit tedious to do, sadly, but thankfully as a minor relief we can stop after three intersections, it is impossible for you to have violated the upper bound on four or more of these variables simultaneously. – JMoravitz May 14 '20 at 00:09
  • 1
    Alternatively, if you just want an answer quickly and aren't afraid of using technology, this is going to be the coefficient of $x^{109}$ in the expansion of $(x^4+x^5+x^6+\dots+x^{29})^2(x^{10}+x^{11}+\dots+x^{40})^2(1+x+x^2+\dots+x^{109})^2$ – JMoravitz May 14 '20 at 00:12
  • @Leavei, be careful with $3<$ versus $3 \le$, and with $>0$ versus nonnegative. – RobPratt May 14 '20 at 01:28

1 Answers1

0

This can be done with generating functions, as suggested by JMoravitz in a comment, and it's quite possible to do it with a pencil, though you'll probably want to use a calculator to do the arithmetic; I certainly did.

First, to understand the comment, note that $x_1$ and $x_2$ are each represented by the polynomial $$x^4+x^5+x^6+\dots+x^{29}.$$ This is because $4\leq x_1,x_2\leq29$. Similarly, $x_3$ and $x_4$ are each represented by the polynomial $$x^{10}+x^{11}+\dots+x^{40}.$$ Finally, $x_5$ and $x_6$ are each represented by the polynomial $$1+x+x^2+\dots+x^{109}$$ Since the sum is to be $109$, neither $x_5$ nor $x_6$ can be more than $109$, and they must be $\geq0$.

To multiply these $6$ polynomials, we choose a term from each polynomial, multiply them together, and add up the products over all choices of terms. The coefficients in the polynomials are all $1$, so the coefficient of $x^n$ in the product, for some natural number $n$, is just the number of ways to pick one term from each polynomial such that the sum of their exponents is $n$. When $n=109$, this is just the solution to our problem. For example, the solution $x_=20,x_2=14,x_3=30,x_4=30,x_5=15,x_6=0$ corresponds to choosing the terms $x^{20},x^{14},x^{30},x^{30},x^{15},1$ from the polynomials, in order.

Now it's just a matter of figuring out the coefficient of $x^{109}$ without multiplying the polynomials.

I'll use the notation $[x^n]p(x)$ to mean the coefficient of $x^n$ in the formal power series $p(x)$. We want $$c=[x^{109}]\left(x^4+\cdots+x^{29}\right)^2 \left(x^{10}+\cdots+x^{40}\right)^2 \left(1+x+x^2+\cdots\right)^2 $$ Note that we don't need an upper bound on the $x_5$ and $x_6$. It doesn't matter if we include exponents $>109$ in the polynomial, because they won't contribute anything to the coefficient of $x^{109}$ in the product. As you'll see, this simplifies the calculation, because we have two fewer factor of the numerator this way.

Then, using the formula for a geometric series,$$ \begin{align} c&=[x^{109}]\left(x^4-x^{30}\right)^2 \left(x^{10}-x^{41}\right)^2(1-x)^{-6}\\ &=[x^{81}]\left(1-x^{26}\right)^2 \left(1-x^{31}\right)^2 \sum_{n=0}^\infty\binom{-6}{n}(-x)^n\\ &=[x^{81}](1-2x^{26}+x^{52})(1-2x^{31}+x^{62}) \sum_{n=0}^{81}(-1)^n\binom{n+5}{5}(-x)^n\\ &=[x^{81}](1-2x^{26}-2x^{31}+x^{52}+4x^{57}+x^{62}) \sum_{n=0}^{81}\binom{n+5}{5}x^n\\ \end{align}$$
since we may ignore terms of degree $>81$.

We just need to pick out the terms in the product that result in a term of degree $81.$ We have $$ \binom{86}{5}-2\binom{60}5-2\binom{55}5+\binom{34}5+4\binom{29}5+\binom{24}5=17,741,536 $$

saulspatz
  • 53,131
  • I'm new to this area of mathematics. How does one solve these type of problems in general? Can you provide some reference? – General Grievous May 30 '20 at 06:36
  • @PankajTiwari I'm not sure what you mean by "these types of problems". – saulspatz May 30 '20 at 06:46
  • @PankajTiwari Do you just mean calculating the number of solutions to equations of this sort, or do you have something broader in mind? – saulspatz May 30 '20 at 06:54
  • I wish to know how to calculate the number of solutions to equations of this short. – General Grievous May 30 '20 at 07:03
  • I wish to know how to calculate the number of solutions to equations of this short. – General Grievous May 30 '20 at 07:03
  • I don't have a reference for you, I'm afraid, but I think this example should show you everything you need. Is there some part of it you don't understand? If so, let me know which and I'll add more detail. It won't be till tomorrow though, probably. I'm about to go to bed. – saulspatz May 30 '20 at 07:05
  • To be honest, I don't understand the whole proof. Can you give a more complete explanation of why and how you used formal power series and the binomial expansion? Thanks. – General Grievous May 30 '20 at 14:26
  • @PankajTiwari Do you understand the very beginning, that is why the coefficient of $x^{109}$ in that particular power series gives the answer? Also, are you familiar with the binomial series from calculus? (I'm trying to see what level of detail I need to add.) – saulspatz May 30 '20 at 15:39
  • I'm aware of the binomial series from calculus but I don't understand why the 109th coefficient gives the answer. – General Grievous May 30 '20 at 16:27
  • @PankajTiwari I added some more explanation. Let me know if anything's still not clear. – saulspatz May 30 '20 at 17:50
  • @PankajTiwari I just answered another question like this, but simpler. I put in a lot of explanation, partly with you in mind. https://math.stackexchange.com/questions/3698795/discrete-mathematics-generating-function/3698943#3698943 – saulspatz May 31 '20 at 00:35
  • Thanks. Things are now clear for me. – General Grievous May 31 '20 at 05:54