Puzzle – terms in the expansion
November 15, 2007 am30 9:10 am
3 Comments
leave one →
Fairly basic, but this is what I do…
How many terms are there in ?
?
How many terms in ?
?
?
?
Can you generalize to more terms? (check yourself with answer is 210 terms.)
For (x1+x2+…xk)^n, I get C(n+k-1,k-1) terms.
This does not agree with your answer for the special case, however. So one of us has made a mistake.
Additionally, this appears to be a special case of the Bose-Einstein statistics problem of counting the number of ways in which n identical balls can be put in k bins, for which the answer above is correct.
Arrgh, you caught me counting too fast. I meant
I think there was an eleven in my mental math (6+5-1), which I dutifully remembered and wrote down, which was wrong.
Thank you.
And yes, in k distinct bins.
Here’s the proof I use (generalizing from an example)
Consider
The five exponents will add up to 6, so we can consider the equivalent problem a+b+c+d+e=6, where a,b,c,d,e are 0 or positive integers.
Represent our exponents as stars: ******
.
Use vertical lines to split them into groups: ***|*||**|
This arrangement corresponds to 3,1,0,2,0 or
Notice that lines without stars in between means that that exponent is 0 (as does a line at the beginning or the end).
All we have to do now is find out how many ways we can rearrange these symbols: C(10,4). Hmm. Where’s the 10 come from? We have 6 exponents, and 6 stars. We have 5 terms, and 4 lines. So exponents + terms – 1 would seem to do it. And the 4? Thats the number of terms – 1.
So for
we will have n stars, k – 1 vertical lines, and our answer is Clueless’: C(n+k-1,k-1).