My Math Forum Probability

 Probability and Statistics Basic Probability and Statistics Math Forum

December 21st, 2018, 06:26 PM   #11
Math Team

Joined: Oct 2011

Posts: 14,597
Thanks: 1038

Quote:
 Originally Posted by romsek seriously? pick the 7 colors, you have 1 left with 7 colors to choose from.
Guess I wasn't too clear!
What I was doing was:
pick 7 only
are they all different color?

1 * 6/55 * 5/54 * 4/53 * 3/52 * 2/51 * 1/50

 December 21st, 2018, 07:44 PM #12 Senior Member     Joined: Jul 2012 From: DFW Area Posts: 642 Thanks: 99 Math Focus: Electrical Engineering Applications With a program, an average of 3 runs of a million trials each gives: Probability of at least 1: 0.069954 Probability of exactly 1: 0.067655 My (perhaps rough) calculation for the probability of the first stack having all different colors - assuming 56 cards, 8 each of 7 colors, and arranged into a stack of 7 cards (per the OP in post #8 ): $\displaystyle \large \frac{56}{56} \cdot \overbrace{\frac{48}{55}}^{\text{ cards of another color left}} \cdot \frac{40}{54} \cdot \frac{32}{53} \cdot \frac{24}{52} \cdot \frac{16}{51} \cdot \frac{8}{50}\approx 0.00904$ Now the following part definitely is rough, and I trust my program results more, but since there are 8 stacks: $\displaystyle \approx 0.00904 \cdot 8 \approx 0.072$ which is in reasonable agreement with my program. I'm unsure how to get the exact number. Edit: Oops had + instead of \cdot Last edited by jks; December 21st, 2018 at 08:01 PM.
 December 31st, 2018, 11:56 AM #13 Newbie   Joined: Dec 2018 From: PA Posts: 5 Thanks: 0 JKS: I got the part you put down with (56/56)*(48/55).... And that works if I'm only making one stack. But once you throw in the 7 other stacks, those top numbers change. But to what, I'm unsure. My reasoning is this. By the time I get back to the first stack, I really don't have 55 cards left. I only have 48...and how many are of a different color from the original card in the first stack is really a variable that will change with each time you put that first card down... I guess I'll have to be content with this "best guess" that you have provided me. It seems like such a high probability, though. Given how many times I've done this (granted, nowhere hear the million your program did three times), I should have had it happen at least once so far.... Thanks for your help! Love and Light, Namaste! Shelly
December 31st, 2018, 12:34 PM   #14
Math Team

Joined: Oct 2011

Posts: 14,597
Thanks: 1038

Quote:
 Originally Posted by jks With a program, an average of 3 runs of a million trials each gives: ............. Now the following part definitely is rough, and I trust my program results more,
56 cards "put in" an array size 56 ?
1st pick at random from 56?
Then 56th array element swapped with element picked?
2nd pick at random from 55?
...and similarly afterwards?

 December 31st, 2018, 07:10 PM #15 Math Team   Joined: Oct 2011 From: Ottawa Ontario, Canada Posts: 14,597 Thanks: 1038 With 1 representing one of the colors, 2 representing another, and so on, then this interesting result (all stacks have the 7 colors) is possible: 1 2 3 4 5 6 7 8 : stacks ----------------- 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 However, probability of hitting that result is 1 in zillions!! Happy new year!
 January 1st, 2019, 11:02 AM #16 Senior Member     Joined: Jul 2012 From: DFW Area Posts: 642 Thanks: 99 Math Focus: Electrical Engineering Applications Happy new year indeed! Since I think that I can actually calculate the probability of all of the stacks containing all colors as in the post above, I will reply in reverse order to the posts. I think that the probability is: \displaystyle \large { \begin{align} & \left (\frac{56}{56} \cdot \frac{48}{55} \cdot \frac{40}{54} \cdot \frac{32}{53} \cdot \frac{24}{52} \cdot \frac{16}{51} \cdot \frac{8}{50} \right ) \\ \cdot & \left (\frac{49}{49} \cdot \frac{42}{48} \cdot \frac{35}{47} \cdot \frac{28}{46} \cdot \frac{21}{45} \cdot \frac{14}{44} \cdot \frac{7}{43} \right ) \\ \cdot & \left (\frac{42}{42} \cdot \frac{36}{41} \cdot \frac{30}{40} \cdot \frac{24}{39} \cdot \frac{18}{38} \cdot \frac{12}{37} \cdot \frac{6}{36} \right ) \\ \cdot & \left (\frac{35}{35} \cdot \frac{30}{34} \cdot \frac{25}{33} \cdot \frac{20}{32} \cdot \frac{15}{31} \cdot \frac{10}{30} \cdot \frac{5}{29} \right ) \\ \cdot & \left (\frac{28}{28} \cdot \frac{24}{27} \cdot \frac{20}{26} \cdot \frac{16}{25} \cdot \frac{12}{24} \cdot \frac{8}{23} \cdot \frac{4}{22} \right ) \\ \cdot & \left (\frac{21}{21} \cdot \frac{18}{20} \cdot \frac{15}{19} \cdot \frac{12}{18} \cdot \frac{9}{17} \cdot \frac{6}{16} \cdot \frac{3}{15} \right ) \\ \cdot & \left (\frac{14}{14} \cdot \frac{12}{13} \cdot \frac{10}{12} \cdot \frac{8}{11} \cdot \frac{6}{10} \cdot \frac{4}{9} \cdot \frac{2}{8} \right ) \\ \cdot & \left (\frac{7}{7} \cdot \frac{6}{6} \cdot \frac{5}{5} \cdot \frac{4}{4} \cdot \frac{3}{3} \cdot \frac{2}{2} \cdot \frac{1}{1} \right ) \\ & = \\ 8^7 & \cdot (7 \cdot 6 \cdot 5 \cdot 4 \cdot 3 \cdot 2 \cdot 1) \\ \cdot 7^7 & \cdot (7 \cdot 6 \cdot 5 \cdot 4 \cdot 3 \cdot 2 \cdot 1) \\ & \vdots \\ \cdot 2^7 & \cdot (7 \cdot 6 \cdot 5 \cdot 4 \cdot 3 \cdot 2 \cdot 1) \\ \cdot 1^7 & \cdot (7 \cdot 6 \cdot 5 \cdot 4 \cdot 3 \cdot 2 \cdot 1) \ / \ 56! \\ & = \frac{(8!)^7(7!)^8}{56!} \approx 1.0 \times 10^{-13} \end{align}} To check, I simplified to 3 stacks instead of 8 and I calculated the probability as: $\displaystyle \large \frac{(3!)^7(7!)^3}{21!} \approx 0.000701$ I also modified my program to 3 stacks and to check only for all stacks containing all colors. An average of 3 runs of 10 million tries gave an average probability of $\approx 0.000699$ so hopefully the analysis is correct. Thanks from romsek
January 1st, 2019, 12:14 PM   #17
Senior Member

Joined: Jul 2012
From: DFW Area

Posts: 642
Thanks: 99

Math Focus: Electrical Engineering Applications
Quote:
 Originally Posted by Denis Curious. On your setup: 56 cards "put in" an array size 56 ? 1st pick at random from 56? Then 56th array element swapped with element picked? 2nd pick at random from 55? ...and similarly afterwards?
You are essentially correct, I just don't do any swapping. First, partly due to the way Ruby works, I built a reference array. For each run I duplicate the reference array into a "cards" array and initialize a "shuffled_cards" array.

I pick a random number 0-55, then add that element to the shuffled array and delete it from the cards array. Then, as you state, I pick another random number from 0-54, and so on.

My code appears below w/o comments but I also attached an image with comments and color coding to make viewing it easier.

Some notes about the program (and Ruby):
• If I say "cards = cards_ref" (both are Arrays) and then make changes to cards, the same changes get made to cards_ref. Hence, .dup is used (line 37 in image).
• # denotes a comment.
• The lines above 26 are just comments w/ the problem statement.
• [] is an empty array and << adds an element to an array.
• Comment out either line 50 or 51. Line 50 makes each stack one at a time, line 51 starts each stack and builds up. As romsek stated in post #5, it should not matter, and from running the program, it does not appear to.
• "Colors" a - g are used, similar to post #6 by Denis, although I could have used anything, of course, including the numbers 1 - 7.
• Ruby obviously makes some things rather easy (such as "letter.next", line 32), but it comes at the expense of speed. I can run about a million tries per minute. C would be much faster.
• For simplicity, the basic program that just looks for at least 1 occurrence of all colors is shown.
• % is the modulo operator.

Here is the code (also see the image):

Code:
num_tries = 1000000
num_all_colors = 0

letter = 'a'
cards_ref = Array.new(8, letter)
6.times do
letter = letter.next
Array.new(8, letter).each{|el| cards_ref << el}
end

num_tries.times do
cards = cards_ref.dup
shuffled_cards = []

for n in 0..55
rand_ind = rand(56 - n)
shuffled_cards << cards[rand_ind]
cards.delete_at(rand_ind)
end

stacks = []
8.times{stacks << []}

for n in 0..55
stacks[n/7][n%7] = shuffled_cards[n]
#  stacks[n%8][n/8] = shuffled_cards[n]
end

all_colors = false
for n in 0..7
if stacks[n].uniq.length == 7
all_colors = true
break
end
end
num_all_colors += 1 if all_colors
end

puts num_tries
puts num_all_colors.to_f / num_tries.to_f
Attached Images
 cards_2018_12_21_prog_img.jpg (95.3 KB, 2 views)

January 1st, 2019, 01:31 PM   #18
Senior Member

Joined: Jul 2012
From: DFW Area

Posts: 642
Thanks: 99

Math Focus: Electrical Engineering Applications
Quote:
 Originally Posted by Rainbowlady JKS: I got the part you put down with (56/56)*(48/55).... And that works if I'm only making one stack. But once you throw in the 7 other stacks, those top numbers change. But to what, I'm unsure. My reasoning is this. By the time I get back to the first stack, I really don't have 55 cards left. I only have 48...and how many are of a different color from the original card in the first stack is really a variable that will change with each time you put that first card down... I guess I'll have to be content with this "best guess" that you have provided me. It seems like such a high probability, though. Given how many times I've done this (granted, nowhere hear the million your program did three times), I should have had it happen at least once so far.... Thanks for your help! Love and Light, Namaste! Shelly
Hi Shelly,

I share your uncertainty of what the numbers should be and for the recognition of the difficulty of accounting for all of the possibilities as different cards are chosen. That is one of the reasons why I decided to try to program in the first place and come up with a rough calculation later (for at least some measure of checking).

Unfortunately, to more precisely calculate the probability, I would have to review some of the probability calculations that I have done in the past (I'm a bit out of practice, and frankly I was never really good at such problems).

I am not sure that I will be able to pursue it much further as work starts again tomorrow. However, if I do, I'm sure that I would start by simplifying to 2 or 3 stacks of 7 cards (similar to what I did with the 'all colors in all stacks' post above) and see if I could make some headway there. If I did, I would then try to extend it to 8, but even this may be difficult (I am unsure).

Of course, there is no reason why you cannot try this for yourself and if needed ask questions along the way. If I cannot help, there are some pretty sharp members who probably can.

Also, for the calculation for a single stack and for the 'all colors in all stacks' I think that you can see that I assumed that the 8 stacks are made one at a time. As romsek pointed out, it should not matter if the stacks are made one at a time or dealt out and built up individually. In my program, I can choose how the stacks are made and it indeed does not seem to matter. So it would be really neat to work the problem both ways and get the same answer, which of course, may be difficult.

Finally, how many times have you run the experiment? I modified my program to count how many tries it took to get at least one stack of all colors and the average was ~14.3. Of course, sometimes it took only 1 try, but sometimes it took over 100.

 Tags probability

 Thread Tools Display Modes Linear Mode

 Similar Threads Thread Thread Starter Forum Replies Last Post Devans99 Probability and Statistics 4 June 25th, 2018 01:06 PM rivaaa Advanced Statistics 5 November 5th, 2015 01:51 PM hbonstrom Applied Math 0 November 17th, 2012 07:11 PM token22 Advanced Statistics 2 April 26th, 2012 03:28 PM naspek Calculus 1 December 15th, 2009 01:18 PM

 Contact - Home - Forums - Cryptocurrency Forum - Top