The Student Room Group

S1 Permutations Combinations

The word DISESTABLISHMENTARIANISM is written out on 24 cards, one on each card. The cards are shuffled and 9 of them are selected and arranged in a straight line.

1. How many possible selections are there of 9 letters?
2. How many arrangements are there of 9 letters?

The first is clearly asking for combinations and the second for permutations but I honestly have no idea how to solve this. It seems some sort of method to cater for nCr and nPr is being requested in tandem with duplication so this will be an instructive problem for me.
(edited 11 years ago)
Original post by Big-Daddy
...


Is this really S1? And have you given the question exactly as worded in the source document?
Reply 2
Original post by ghostwalker
Is this really S1? And have you given the question exactly as worded in the source document?


Yes to both. It's an S1 extension question my teacher said we needed to figure out how to do.

I'm aware that this requires both nCr and nPr along with duplication modifications. I just don't know how to do that yet.
Original post by Big-Daddy
Yes to both. It's an S1 extension question my teacher said we needed to figure out how to do.

I'm aware that this requires both nCr and nPr along with duplication modifications. I just don't know how to do that yet.


Never heard of S1 extension.

I'd start by counting the number of each letter.

Then one method would be to break it down into different cases depending on the duplication factors.

E.g no. of choices such that you have 4 of one letter, 4 of another, and 1 of something else.

By my reckoning there are 18 cases to consider, and some of them quite complex to evaluate. You'll need to be methodical.

Rather tedious, so makes me wonder if there is a simpler method; perhaps involving polynomials?
Reply 4
Original post by ghostwalker
Never heard of S1 extension.

I'd start by counting the number of each letter.

Then one method would be to break it down into different cases depending on the duplication factors.

E.g no. of choices such that you have 4 of one letter, 4 of another, and 1 of something else.

By my reckoning there are 18 cases to consider, and some of them quite complex to evaluate. You'll need to be methodical.

Rather tedious, so makes me wonder if there is a simpler method; perhaps involving polynomials?


I can only imagine that the reason my teacher would set an example like "DISESTABLISHMENTARIANISM" - famous for its length - would be to discourage the approach which involves tabulating everything and going through it all bit by bit. So there is probably a more rigorous solution, though it may be somewhat more complex in terms of the maths.

How about this for the combinations:

n=24, r=9, d=3! (I) * 4! (S) * 2! (E) * 2! (T) * 3! (A) * 2! (M) * 2! (N)

And then just plug into my general formula (I have no clue whether it will work for this, though it did solve a simpler example I looked at earlier today):

N=((n!)/(d!))/(nCr) where d! is the correction for repetition.
N=3.4326*1013.

So would you say my general formula is correct?
Original post by Big-Daddy

How about this for the combinations:

n=24, r=9, d=3! (I) * 4! (S) * 2! (E) * 2! (T) * 3! (A) * 2! (M) * 2! (N)

And then just plug into my general formula (I have no clue whether it will work for this, though it did solve a simpler example I looked at earlier today):

N=((n!)/(d!))/(nCr) where d! is the correction for repetition.
N=3.4326*1013.

So would you say my general formula is correct?


Don't follow what you're doing there.

However, if all the letters were different there would be 24C9 = 1307504 combinations. Since there is duplication the actual number will be less than that, so what you have can't be right.

Edit: Looking at it in terms of polynomials, if I got it right,

Spoiler

(edited 11 years ago)
Reply 6
Original post by ghostwalker
Don't follow what you're doing there.

However, if all the letters were different there would be 24C9 = 1307504 combinations. Since there is duplication the actual number will be less than that, so what you have can't be right.

Edit: Looking at it in terms of polynomials, if I got it right,

Spoiler



OK, it's probably better we leave that method aside.

How would I do it if rather than finding combinations of 9 letters I had to find combinations of all 24? (Given that there is some duplication.) That is easier I'm sure. e.g. does n just fall to the number of distinct elements and then (n-d)Cr=(n-d)C(n-d)=1?
(edited 11 years ago)
Original post by Big-Daddy
OK, it's probably better we leave that method aside.

How would I do it if rather than finding combinations of 9 letters I had to find combinations of all 24? (Given that there is some duplication.) That is easier I'm sure. e.g. does n just fall to the number of distinct elements and then (n-d)Cr=(n-d)C(n-d)=1?


If you want combinations of all 24, then yes, there's only 1.

But, n doesn't fall to the number of distinct elements. I wouldn't even use combinatorics for it. There can be only one way to choose 24 letters from 24.

With some degree of duplication things get quite complex, and I can't see an easy way to do the original question without a lot of work; doesn't mean such a method doesn't exist mind.
(edited 11 years ago)

Quick Reply

Latest