Nested sum of units

What is the value of n n such that a 1 = 0 4 a 2 = 0 a 1 a n = 0 a n 1 1 = 1001 ? \displaystyle {\large \sum_{a_1=0}^4 \sum_{a_2=0}^{a_1} \cdots \sum_{a_n=0}^{a_{n-1}} 1} = 1001 ?


The answer is 10.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

5 solutions

Brian Moehring
Sep 23, 2018

Note that a 1 = 0 4 a 2 = 0 a 1 a n = 0 a n 1 1 = 0 a n a n 1 a 1 4 1 = { ( a 1 , a 2 , , a n ) Z n : 0 a n a 2 a 1 4 } = ( n + 5 1 5 1 ) (Using Stars and Bars) = ( n + 4 4 ) \begin{aligned} \large \sum_{a_1=0}^4 \sum_{a_2=0}^{a_1} \cdots \sum_{a_n=0}^{a_{n-1}} 1 &= \large \sum_{0 \leq a_n \leq a_{n-1} \leq \cdots \leq a_1 \leq 4} 1 \\ &= \Bigg|\{(a_1,a_2,\ldots,a_n) \in \mathbb{Z}^n : 0 \leq a_n \leq \cdots \leq a_2 \leq a_1 \leq 4\}\Bigg| \\ &= \binom{n+5-1}{5-1} & \text{(Using Stars and Bars)}\\ &= \binom{n+4}{4} \end{aligned}

Then ( n + 4 4 ) = 1001 n = 10 \binom{n+4}{4} = 1001 \implies n=\boxed{10}


Remark : It looks from the comments that there's some confusion about how to use stars-and-bars to count this. Here's an illustration I've given to try to explain it: (note that all "left" and "right" directions are from an outside, fixed observer)

Put n + 4 n+4 people standing side-by-side in a line. Give four of them chairs and have them sit down. Then tell the remaining n n people to count how many people to their right are sitting. The number counted by the k k th-from-the-left standing person is a k . a_k.

This defines a bijection between the number of ways to choose 4 4 people from n + 4 n+4 and the number of [non-strictly] decreasing sequences of length n n from { 0 , 1 , 2 , 3 , 4 } \{0,1,2,3,4\}

Sir i don't understand

Sun Tzu - 2 years, 8 months ago

Log in to reply

Yeah me neither

Anuj Tripathi - 2 years, 8 months ago

Inevitably someone doesn't understand something , but if you're not merely stating that truism, you need to meet me halfway. Which part of the solution don't you understand?

  • The notation used in the problem? If so, which of the following: n , , a n , a 2 n, \sum, a_n, a_2 etc.
  • The notation used in my solution? If so, which of the following: , 0 a n a 1 4 , , { } , Z n , , ( m k ) \leq, {\displaystyle \sum_{0\leq a_n \leq \cdots \leq a_1 \leq 4}}, |\cdot|, \{\cdots\}, \mathbb{Z}^n, \in, \binom{m}{k} etc
  • Any single expression in my solution? If so, which one?
  • Why any single step in the solution is true? If so, which ' = = ' is the step you don't understand?
  • Why I solved it this way?

For the record, I have honestly run into people who didn't understand something in each of those categories. I estimate I spend about 90% of the time responding to "I don't understand" trying to figure out what the person doesn't understand.

Brian Moehring - 2 years, 8 months ago

If I have correctly interpreted the solution, Mr Moehring argues that the desired sum is equal to the number of sets { a 1 , a 2 , , a n } \{a_{1}, a_{2}, \cdots , a_{n}\} such that 0 a n a n 1 a 2 a 1 4 0 \leq a_{n} \leq a_{n-1} \leq \cdots \leq a_{2} \leq a_{1} \leq 4 . He counts the number of such sets using technique called Stars and Bars.

Consider the sequence of 4 stars and n n bars, where the total number of stars before i t h i^{th} bar equals a n + 1 i a_{n+1-i} . Notice how this setting nicely disallows a i < a i 1 a_{i} < a_{i-1} . For example, if n = 3 n = 3 , the sequence \star|\star\star |\,|\star implies a 1 = 3 , a 2 = 3 , a 3 = 1 a_{1} = 3, \, a_{2} = 3, \, a_{3} = 1 . Thus, the problem now breaks down to counting the number of ways of arranging 4 stars and n n bars, which is equal to ( n + 4 4 ) \binom{n+4}{4} .

Uros Stojkovic - 2 years, 8 months ago

Log in to reply

Other than a technical point due to how we define "set" (we need to count repetition and sets don't allow that), this is all correct.

Brian Moehring - 2 years, 8 months ago

Log in to reply

@Brian Moehring Agree, tuple is more technically correct term here.

Uros Stojkovic - 2 years, 8 months ago

I feel unknown to this study of maths. I am not that bad at maths though. Tgrow aome light.

Karan Chopra - 2 years, 8 months ago

Could you please clarify in your solution why the cardinality of the set is the same as the number of ways to distribute 4 4 objects to n + 1 n+1 people and not to n n people. Of course experienced problem solvers can make this connection themselves, but some might be unable to make this connection even if they know the stars and bars method.

Jesse Nieminen - 2 years, 8 months ago

For whatever reason, I can no longer post a response directly to @Jesse Nieminen . This is what I wanted to say:

Put n + 4 n+4 people standing side-by-side in a line. Give four of them chairs and have them sit down. Then tell the remaining n n people to count how many people to their right are sitting. The number counted by the k k th-from-the-left standing person is a k . a_k.

This defines a bijection between the number of ways to choose 4 4 people from n + 4 n+4 and the number of [non-strictly] decreasing sequences of length n n from { 0 , 1 , 2 , 3 , 4 } \{0,1,2,3,4\}

Brian Moehring - 2 years, 8 months ago

Log in to reply

This is a more intuitive method than using the stars and bars method, nice.

Could you edit your solution and replace the usage of the stars and bars method with this one? The reason I wrote that previous comment is still valid, connection between the number of these sequences to the conditions for the stars and bars method might be difficult to grasp.

Jesse Nieminen - 2 years, 8 months ago

Log in to reply

This actually is stars-and-bars. I'm putting n n people into 5 5 groups corresponding to their numbers by separating them with 4 4 "bars" (the people in the chairs).

I'll add the illustration to a remark at the end of the solution, though, if you think it's helpful.

Brian Moehring - 2 years, 8 months ago

While the solution to the final part can be found by trial and error, it can also be discovered if we know that 1001= 7 x 11 x 13.

We are looking for n s.t. (n+4)(n+3)(n+2)(n+1)/24 = 7 x 11 x 13 and the product of the sequence of numbers {11,12,13,14} fits the bill perfectly

Malcolm Rich - 2 years, 8 months ago

Although I initially didn't understand the notation for a summation over sets, the comments cleared up my confusion. Very insightful solution. I find both the stars and bars explanation and the bijection explanation useful for understanding how to finish the problem.

Chethan Bhateja - 2 years, 8 months ago
X X
Sep 18, 2018

a 1 = 0 4 1 = 1 + 1 + 1 + 1 + 1 = 5 = ( 5 4 ) \sum_{a_1=0}^4 1=1+1+1+1+1=5=\binom{5}{4} a 1 = 0 4 a 2 = 0 a 1 1 = 1 + 2 + 3 + 4 + 5 = 15 = ( 6 4 ) \sum_{a_1=0}^4\sum_{a_2=0}^{a_1} 1=1+2+3+4+5=15=\binom{6}{4} a 1 = 0 4 a 2 = 0 a 1 a 3 = 0 a 2 1 = 1 + 3 + 6 + 10 + 15 = 35 = ( 7 4 ) \sum_{a_1=0}^4\sum_{a_2=0}^{a_1}\sum_{a_3=0}^{a_2} 1=1+3+6+10+15=35=\binom{7}{4}

\vdots a 1 = 0 4 a 2 = 0 a 1 a n = 0 a n 1 1 = ( 4 + n 4 ) = 1001 \sum_{a_1=0}^4 \sum_{a_2=0}^{a_1} \cdots \sum_{a_n=0}^{a_{n-1}} 1 = \binom{4+n}{4}=1001 So, n = 10 n=10


Proof: a 1 = 0 m 1 = m + 1 = ( m + 1 1 ) \sum_{a_1=0}^m 1=m+1=\binom{m+1}{1}

a 1 = 0 m a 2 = 0 a 1 1 = a 1 = 0 m ( a 1 + 1 1 ) = ( m + 2 2 ) \sum_{a_1=0}^m\sum_{a_2=0}^{a_1} 1=\sum_{a_1=0}^m \binom{a_1+1}{1}=\binom{m+2}{2}

a 1 = 0 m a 2 = 0 a 1 a 3 = 0 a 2 1 = a 1 = 0 m ( a 1 + 2 2 ) = ( m + 3 3 ) \sum_{a_1=0}^m\sum_{a_2=0}^{a_1}\sum_{a_3=0}^{a_2} 1=\sum_{a_1=0}^m \binom{a_1+2}{2}=\binom{m+3}{3}

a 1 = 0 4 a 2 = 0 a 1 a 3 = 0 a 2 a 4 = 0 a 3 1 = a 1 = 0 m ( a 1 + 3 3 ) = ( m + 4 4 ) \sum_{a_1=0}^4\sum_{a_2=0}^{a_1}\sum_{a_3=0}^{a_2}\sum_{a_4=0}^{a_3} 1=\sum_{a_1=0}^{m}\binom{a_1+3}{3}=\binom{m+4}{4}

\vdots

a 1 = 0 m a 2 = 0 a 1 a n = 0 a n 1 1 = a 1 = 0 m ( a 1 + ( n 1 ) n 1 ) = ( m + n n ) \sum_{a_1=0}^m \sum_{a_2=0}^{a_1} \cdots \sum_{a_n=0}^{a_{n-1}} 1=\sum_{a_1=0}^m\binom{a_1+(n-1)}{n-1}=\binom{m+n}{n}

Put in m = 4 m=4 will get the above result.

It's good that you have spotted the pattern. For completeness, can you prove that a 1 = 0 4 a 2 = 0 a 1 a n = 0 a n 1 1 = ( 4 + n 4 ) \displaystyle \sum_{a_1=0}^4 \sum_{a_2=0}^{a_1} \cdots \sum_{a_n=0}^{a_{n-1}} 1 = \binom{4+n}{4} ?

Pi Han Goh - 2 years, 8 months ago

Log in to reply

I'll write a better proof later. Maybe I'll use something about the Pascal's Triangle

X X - 2 years, 8 months ago
Jeremy Galvagni
Sep 23, 2018

I needed to reason this out without all those pesky Sigmas.

If n=0 it's just a single 1 with no summation. 1 1

If n=1 it's just the number 1 added to itself 5 times. 1 + 1 + 1 + 1 + 1 = 5 1+1+1+1+1=5

If n=2 we need to make the next summation do each of 0,1,2,3,4 and so the sum is ( 1 ) + ( 1 + 1 ) + ( 1 + 1 + 1 ) + ( 1 + 1 + 1 + 1 ) + ( 1 + 1 + 1 + 1 + 1 ) = 1 + 2 + 3 + 4 + 5 = 15 (1)+(1+1)+(1+1+1)+(1+1+1+1)+(1+1+1+1+1)=1+2+3+4+5=15

If n=3 we need to make the next summation do each of 0,1,2,3,4 on the previous so the sum is ( 1 ) + ( 1 + 2 ) + ( 1 + 2 + 3 ) + ( 1 + 2 + 3 + 4 ) + ( 1 + 2 + 3 + 4 + 5 ) = 1 + 3 + 6 + 10 + 15 = 35 (1)+(1+2)+(1+2+3)+(1+2+3+4)+(1+2+3+4+5)=1+3+6+10+15=35

Adding numbers like this is one way of using Pascal's Triangle: the Hockey Stick Identity

We want hockey sticks 5 numbers long. This means we want the 5th number in each row (counting starts with 0)

A glance at the Triangle shows ( 14 4 ) = 1001 \binom{14}{4}=1001 but since n=1 is in row 5: 4 numbers ahead, the solution is n = 14 4 = 10 n=14-4=\boxed{10}

Very easy to understand. Thanks for the post!

Chethan Bhateja - 2 years, 8 months ago
Uros Stojkovic
Sep 24, 2018

Define f ( x , n ) = a 1 = 0 x a 2 = 0 a 1 a n = 0 a n 1 1 \displaystyle{f(x,n) = \sum_{a_{1} = 0}^{x}\sum_{a_{2} = 0}^{a_{1}}\cdots\sum_{a_{n} = 0}^{a_{n-1}}1} . Immediately it follows that f ( 0 , n ) = 1 f(0,n) = 1 and f ( x , n ) = 0 f(x,n) = 0 for all x < 0 x < 0 . Moreover, we have: f ( x , n ) = f ( x 1 , n ) + f ( x , n 1 ) . f(x,n) = f(x-1,n)+f(x,n-1). To keep the above relation consistent, set f ( x , 0 ) = 1 f(x, 0) = 1 . Consider this sequence: 1 f ( x , n ) = 1 f ( x 1 , n ) + 1 f ( x , n 1 ) = 1 f ( x 2 , n ) + 2 f ( x 1 , n 1 ) + 1 f ( x , n 2 ) = 1 f ( x 3 , n ) + 3 f ( x 2 , n 1 ) + 3 f ( x 1 , n 2 ) + 1 f ( x , n 3 ) = \begin{aligned} {\color{#3D99F6} 1}f(x&,n)= \\ {\color{#3D99F6} 1}f(x-1,n)\, &+\, {\color{#3D99F6} 1}f(x,n-1)= \\ {\color{#3D99F6} 1}f(x-2,n)\, +\, {\color{#3D99F6} 2}f(x-1&,n-1)\, +\, {\color{#3D99F6} 1}f(x,n-2)= \\ {\color{#3D99F6} 1}f(x-3,n)\, +\, {\color{#3D99F6} 3}f(x-2,n-1)\, &+\, {\color{#3D99F6} 3}f(x-1,n-2)\, +\, {\color{#3D99F6} 1}f(x,n-3) = \\ &~\vdots\end{aligned} One can easily recognize that coefficients correspond to Pascal triangle. Next, notice that coefficient of f ( x a , n b ) f(x-a, n-b) is equal to ( a + b a ) = ( a + b b ) \displaystyle{\binom{a+b}{a} = \binom{a+b}{b}} .

Further on, deduce that the whole sum is going to converge to one element of our Pascal triangle, namely the one corresponding to f ( 0 , 0 ) f(0,0) : f ( x , n ) = ( x + n x ) f ( 0 , 0 ) = ( x + n x ) 1 f(x, n) = \binom{x + n}{x} \cdot f(0,0) = \binom{x+n}{x} \cdot 1

Finally, plugging in values x = 4 x = 4 and f ( 4 , n ) = 1001 f(4, n) = 1001 yields n = 10. n = 10.

Oh god, this is beautiful. I was trying to come up with a recursive method, but I don't know why I didn't recognize the hidden Pascal triangle, even though I clearly know the multi-summation can be expressed as a binomial coefficient.

Thank you for this alternative solution.

Pi Han Goh - 2 years, 8 months ago

Log in to reply

Thanks for kind words. Brian's solution is also great.

Uros Stojkovic - 2 years, 8 months ago
Vinod Kumar
Sep 25, 2018

Finish 9 sums at the right end of the expression and simplify it as:

[(j+9)!/{(j!)(9!)}].

After 1 more explicit summation and using WolframAlpha get:

[sum_(j=0,4){((j + 9)!)/(9!)(j!)}] = 1001

Answer=10

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...