stat272 assignment help! (1 Viewer)

[jake2.0]

Hey guys, just need a little help with the 3rd stat272 assignment help. Its on the questions regarding the 'generate random numbers...' questions.

For Q1 i) I said something like get the uniform random number (e.g. 7) and divide it by 10 (giving 0.7) and then find the corresponding Z value from the standard normal table (0.7 gives a Z of roughly 0.53). This is a normal random number so take the exp{} of it to give a lognormal random number. Any of this right???

As for Q2 (h) and (i) i have no idea so if anybody could help it'd be greatly appreciated.

Thanks!

 

[hello2004]

If you post the question I might be able to help.

I think you're on the right track but it looks like you've calculated a two sided probability. You need to calculate a one sided probability since the CDF is P(X<=x).

 

[jake2.0]

If you post the question I might be able to help.

I think you're on the right track but it looks like you've calculated a two sided probability. You need to calculate a one sided probability since the CDF is P(X<=x).

Here is Q1 and Q2

Thanks

 

[hello2004]

Now that I see it I think they want you to use the central limit theorem (not the inverse CDF method) So:

- add up a whole bunch of numbers from the table (say 20)
- divide each number by 100,000 (I'm assuming that table is ~ U(0,100000) )
- The 20 numbers are now ~ U(0,1). Add them up and divide by 20.
- subtract 0.5 from this . This is E(Xbar)
- divide by SQRT[(1/12)/20] This is the StandardDeviation(Xbar)
- You now have a variable which is approximately standard normal

- Exponentiate it !

 

[dracover]

is a bit late for me to add my 2 bobs now but y would u add 20 numbers and divide by 20??? by ur logic if i had an endlessly long number of random numbers then did the whole thing i would always get the mean because thats wat ur effectively doing rather then just a random number

after u get ur U(0,1) just let it equal the cdf of a N(0,1) find the inverse and sub it in. u get ur random normal then exponentiate


part h) its just ur usual generate random numbers isnt it? get a U(0,1) then find let this = P(z) and find the inverse u have ur number

i) z is a -log so if u got 10 thousand no. from it u would get a -log shape but discrete of course...so a kind downward curve from 0-1

 

[jake2.0]

Does anyone have the answers to last years stat271 assignment 1?

Thanks