Magicwand started Masters degree in Operations Research

[DrStraw]

I will ask my professor to make everyone feel better

If he has not already told you that it is done on the honor system then he is very lax. I always have it in big letters at the top of the test.

[Magicwand]

I am trying to build excel plot of Cumulative normal probability vs residuals.
i got residuals of the data but how do i get probability of each residual data?

I am suppose to work it with minitab that automatically draw the graph...
but i am trying to understand instead of using mini-tab.

can someone give me help on how i get prob of each residual?

Edit:
Solved it!
Standarize it and use normsdist() function to get percentage...
thank you.

edit:
but graph i get is different from the graph on the answer book
back to square 1

[wineandgolover]

Umpteen years later and they still use minitab. Makes my day.

[Magicwand]

Umpteen years later and they still use minitab. Makes my day.

i remember 23 or 22 years ago i used it in actuarial class 121 in university of toranto.
my impression of minitab is that excel can do same functions and why anyone want to pay for minitab?

[Magicwand]

i think actual file will help others help me better...
i have attached week 4 homework excel spread sheet.
now my question is why is my risidual vs normal prob curve different from answer book?
someone please help me because i can not figure this out.

[DrStraw]

i think actual file will help others help me better...
i have attached week 4 homework excel spread sheet.
now my question is why is my risidual vs normal prob curve different from answer book?
someone please help me because i can not figure this out.

It would be nice to know what the problem was actually saying and asking. All I saw was a much of data and graphs.

[Magicwand]

Yes it would be helpful
pink section is last week HW i did
this week i am using same data to
a)calculate R^2 which i got same answer as answerbook.
b) Prepare a normal probability plot of the residuals from the least squares model.
c) Plot the risiduals versus the fitted values and against X.
i got c) similar to answer book.
problem is b)
i standarized residual by dividing it by square root of sigma square.
then i used normsdist function in excel to get percentage of standarized residual d.
i plotted d vs percentage.
but the plot appears to be different.

what did i do wrong?

[drmwc]

I've also got a PhD in pure maths, altough I don't work anywhere near academia at the moment.

Was your chart for 11-53(b) similar to the book solution? You've used the same technique for both, so it would seem odd if one is right and the other wrong. The chart technique seems like a reasonable heuristic for showing normality (or not, as the case may be).

What did the book answer look like?

[Magicwand]

I've also got a PhD in pure maths, altough I don't work anywhere near academia at the moment.

Was your chart for 11-53(b) similar to the book solution? You've used the same technique for both, so it would seem odd if one is right and the other wrong. The chart technique seems like a reasonable heuristic for showing normality (or not, as the case may be).

What did the book answer look like?

what other students did was to use mini-tab to draw which was what answer book had.
i used what little math i know and used excel to draw the plot.
book answer was similar but didnt have funny curve that i had.
i can wait for my professor to reply to my homework... he will only decuct few points so really dont matter..
other than that plot i think i had everything else correct i think.

thank you for trying.

[Magicwand]

Problem 1
Each month, a gas station sells 4,000 gallons of gasoline. Each time the parent company refills the stations's tanks, it charges the station $50 plus 70¢, per gallon. The annual cost of holding a gallon of gasoline is 30¢.
We are given that K = $50, h = 30¢/gallon/year, and D = 48,000 gallons/ year.
a. How large should the station's order be?
The EOQ is q* = (2KD/h)1/2
= (2(50)(48000)/.3)1/2
= 4000

b. How many order per year will be placed?
Order/Year = D/q* = 48000/4000 = 12

c. How long will it be between orders?
Length of a cycle = q*/D = 4000/48000 = 1/12 = one month

d. Would the EOQ Assumptions be satisfied in this situation? Why or Why not?
EOQ Assumption:
1. Demand is deterministic and occurs at a constant rate.
2. If an order of any size (say, q units) is place, an ordering and setup cost K is incurred.
3. The lead time for each order is zero.
4. No shortages are allowed.
5. The cost per unit-year of holding inventory is h
In real situation such as above demand will not be constand and lead time can not be zero.
Lead time not being zero also means that shortage may occur.

e. If the lead time is two weeks, what is the reorder point? If the lead time is ten weeks, what is the reorder point? Assume 1 week = 1/52 year.

For a 2-week lead time, reorder point = 48,000 x 2/52 = 1,846.15 gallons.
For a 10-week lead time, reorder point = 1,230.77 gallons.

can someone explain how leadtime of 10 week ordering point is less that 2 week ordering point?
answerbook said following:
If Leadtime = 10 weeks, then Leadtime demand > EOQ. Assume that an order arrives at time T=0. then an order must have been placed at T = -10. We know that an order is placed every 52/12 = 13/3 weeks. thus an order was place at T = 0 we will have 4000 gallons in stock, we know that at the reorder point T = 3 (and any other reorder point) we will have
4000-(3/52)48,000 = 1231 in stock.

reading that paragraph many times and can not understand what they said.

thanks in advance.

[tj86430]

can someone explain how leadtime of 10 week ordering point is less that 2 week ordering point?
answerbook said following:
If Leadtime = 10 weeks, then Leadtime demand > EOQ. Assume that an order arrives at time T=0. then an order must have been placed at T = -10. We know that an order is placed every 52/12 = 13/3 weeks. thus an order was place at T = 0 we will have 4000 gallons in stock, we know that at the reorder point T = 3 (and any other reorder point) we will have
4000-(3/52)48,000 = 1231 in stock.

reading that paragraph many times and can not understand what they said.

thanks in advance.

Which part you don't understand? What it says is that you have be over two cycles ahead in your reorders, because the lead time is so long. And because of that, you can have less left than with a lead time of two weeks.

Another way of putting it is this: reorder point = (48000 x lead time in weeks / 52) Modulus 4000.

lead time of two gives (48000 x 2 / 52) Modulus 4000 = 1846.15
lead time of ten gives (48000 x 10 / 52) Modulus 4000 = 1230.77

Similarly a lead time of twenty would give 2461.54 etc

[Magicwand]

Which part you don't understand? What it says is that you have be over two cycles ahead in your reorders, because the lead time is so long. And because of that, you can have less left than with a lead time of two weeks.

Another way of putting it is this: reorder point = (48000 x lead time in weeks / 52) Modulus 4000.

lead time of two gives (48000 x 2 / 52) Modulus 4000 = 1846.15
lead time of ten gives (48000 x 10 / 52) Modulus 4000 = 1230.77

Similarly a lead time of twenty would give 2461.54 etc

understood ..thank you.
way you explain is much simpler than the answerbook.

[Magicwand]

A drug store sells 30 bottles of antibiotics per week. Each time it orders antibiotics, there is a fixed ordering cost of $10 and a cost of $10/bottle. Assume that the annual holding cost is 20% of the cost of a bottle of antibiotics, and suppose antibiotics spoil and cannot be sold if they spend more than one week in inventory. When the drug store places an order, how many bottles of antibiotics should be ordered?
My answer is : If antibiotics spoil in one week and they sell 30 bottles per week only solution is to order 30 bottles everyweek. but not comfortable with this answer. can anyone do better?

[skydyr]

A drug store sells 30 bottles of antibiotics per week. Each time it orders antibiotics, there is a fixed ordering cost of $10 and a cost of $10/bottle. Assume that the annual holding cost is 20% of the cost of a bottle of antibiotics, and suppose antibiotics spoil and cannot be sold if they spend more than one week in inventory. When the drug store places an order, how many bottles of antibiotics should be ordered?
My answer is : If antibiotics spoil in one week and they sell 30 bottles per week only solution is to order 30 bottles everyweek. but not comfortable with this answer. can anyone do better?

I can't help but wonder if there's a typo and they meant to say that the antibiotics spoil after one month or year or something.

[Magicwand]

A drug store sells 30 bottles of antibiotics per week. Each time it orders antibiotics, there is a fixed ordering cost of $10 and a cost of $10/bottle. Assume that the annual holding cost is 20% of the cost of a bottle of antibiotics, and suppose antibiotics spoil and cannot be sold if they spend more than one week in inventory. When the drug store places an order, how many bottles of antibiotics should be ordered?
My answer is : If antibiotics spoil in one week and they sell 30 bottles per week only solution is to order 30 bottles everyweek. but not comfortable with this answer. can anyone do better?

I can't help but wonder if there's a typo and they meant to say that the antibiotics spoil after one month or year or something.

It is good to know that i am not the only one who think that way.