|
Search: id:A111097
|
|
|
| A111097 |
|
Maximum likelihood estimate of the number of distinguishable marbles in an urn if repeated random sampling of one marble with replacement yields n different marbles before the first repeated marble. |
|
+0
1
|
|
| 1, 2, 5, 8, 13, 19, 25, 33, 42, 51, 62, 74, 86, 100, 115, 130, 147, 165, 183, 203, 224, 245, 268, 292, 316, 342, 369, 396, 425
(list; graph; listen)
|
|
|
OFFSET
|
1,2
|
|
|
COMMENT
|
The numbers in my sequence match the first 8 nonzero numbers in sequence A083704 and the first 40 terms (at least) remain very close to those in A083704 . The program below is very crude and will yield the first 20 terms. To get further terms, one must increase the maximum value of n and then increase the maximum value of k so that the maximum value of k is larger than a(n) for the maximum value of n.
|
|
FORMULA
|
No formula yet, but I want to stake my claim.
|
|
EXAMPLE
|
a(3)=5 because of all urns containing marbles numbered 1, 2, 3, ..., k, an urn containing k = 5 marbles has the largest probability of yielding 3 different marbles before a first repeated marble when sampling with replacement.
|
|
MATHEMATICA
|
thetable = Table[N[n/k*Product[(k - i)/k, {i, 1, n - 1}]], {n, 1, 20}, {k, 1, 300}]; maximums = Map[Max, thetable]; maximumlikelihoodestimates = {}; For[i = 1, i <= Length[thetable], i++, maximumlikelihoodestimates = Append[maximumlikelihoodestimates, Position[thetable[[i]], maximums[[i]]]]]; maximumlikelihoodestimates
|
|
CROSSREFS
|
Cf. A027916. [From R. J. Mathar (mathar(AT)strw.leidenuniv.nl), Aug 18 2008]
Sequence in context: A000982 A122221 A083704 this_sequence A027916 A054254 A025216
Adjacent sequences: A111094 A111095 A111096 this_sequence A111098 A111099 A111100
|
|
KEYWORD
|
nonn
|
|
AUTHOR
|
Marc A. Brodie (mbrodie(AT)wju.edu), Oct 13 2005
|
|
|
Search completed in 0.002 seconds
|
