Timeline for How to generate a RandomVariate of a custom distribution?
Current License: CC BY-SA 3.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 4, 2012 at 13:39 | answer | added | wolfies | timeline score: 5 | |
| Nov 2, 2012 at 9:51 | comment | added | chris | @Rojo old is beautiful ;-) | |
| Nov 2, 2012 at 9:49 | comment | added | chris | @J.M. I have: mathematica.stackexchange.com/questions/14009/… | |
| Nov 2, 2012 at 1:43 | history | tweeted | twitter.com/#!/StackMma/status/264180971115524097 | ||
| Nov 2, 2012 at 1:35 | comment | added | J. M.'s missing motivation♦ | @chris, sounds like a new question... | |
| Nov 2, 2012 at 1:32 | comment | added | J. M.'s missing motivation♦ |
@Sasha, think of Release[] as what Mathematica had before they split up that function into ReleaseHold[] and Evaluate[]. A chimera, if you will...
|
|
| Nov 2, 2012 at 0:48 | comment | added | Rojo | @Sasha I think you just learned something old | |
| Nov 1, 2012 at 21:14 | comment | added | chris |
@Sasha ah! that's just because I learned mathematica 1.2 which had Release which has been depreciated since
|
|
| Nov 1, 2012 at 20:56 | answer | added | Sjoerd C. de Vries | timeline score: 12 | |
| Nov 1, 2012 at 20:46 | comment | added | Sasha |
@chris I did not know of existence of Release. The current way is to use Evaluate. The kudos was because I learned something new.
|
|
| Nov 1, 2012 at 19:51 | comment | added | Sasha |
@chris Kudos for Release!
|
|
| Nov 1, 2012 at 19:40 | comment | added | chris |
@J.M. and more importantly why is it wrong?? Show[Plot[PDF[\[ScriptCapitalD], x], {x, -5, 5}], RandomVariate[\[ScriptCapitalD], 1500]//Histogram[#, Automatic, "Probability"]&]
|
|
| Nov 1, 2012 at 19:35 | comment | added | chris |
@J.M. why does RandomVariate[\[ScriptCapitalD], 15] from reference.wolfram.com/mathematica/ref/… works though?
|
|
| Nov 1, 2012 at 18:39 | history | edited | J. M.'s missing motivation♦ | CC BY-SA 3.0 |
deleted 37 characters in body
|
| Nov 1, 2012 at 16:41 | comment | added | J. M.'s missing motivation♦ |
FWIW: it looks to me (after some experimentation) that RandomVariate[] is not equipped to handle arbitrary discrete probability distributions; in general, you might have to roll your own algorithm.
|
|
| Nov 1, 2012 at 16:41 | review | First posts | |||
| Nov 1, 2012 at 19:17 | |||||
| Nov 1, 2012 at 16:37 | answer | added | chris | timeline score: 19 | |
| Nov 1, 2012 at 16:32 | comment | added | J. M.'s missing motivation♦ |
"the distribution I want is a derived mixture of two Poisson distributions" - could you maybe specify how the two Poisson distributions are mixed? I presume you've tried MixtureDistribution[] already?
|
|
| Nov 1, 2012 at 16:28 | comment | added | chris |
It would work if for instance your new distribution is derived from a known one via TransformedDistribution as in RandomVariate[ TransformedDistribution[x - 1, x \[Distributed] NormalDistribution[]] // Release, 15]
|
|
| Nov 1, 2012 at 16:24 | history | asked | Rafal | CC BY-SA 3.0 |