Timeline for How does the formula for generating correlated random variables work?
Current License: CC BY-SA 3.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 23, 2015 at 0:46 | answer | added | Dmitry Rubanovich | timeline score: 5 | |
| Mar 13, 2015 at 18:24 | history | tweeted | twitter.com/#!/StackStats/status/576448753051312130 | ||
| Mar 12, 2015 at 23:51 | comment | added | whuber♦ | @Anthony The question only asks about correlation, which is purely a function of first and second moments. The answer does not depend on any other properties of the distributions. What you are discussing is a different subject altogether. | |
| S Mar 12, 2015 at 22:17 | history | suggested | D.W. | CC BY-SA 3.0 |
Clean up the question and title. Ask only one question (one question per question); the other question can be asked separately.
|
| Mar 12, 2015 at 22:14 | review | Suggested edits | |||
| S Mar 12, 2015 at 22:17 | |||||
| Mar 12, 2015 at 17:25 | comment | added | Anthony | Very interesting link. I'm not sure I understand what you mean by normality being irrelevant. If $X_1$ or $X_2$ is not normal, and it becomes harder to control the density of $Y$ through the Kaiser-Dickman algorithm. This is the whole reason for specialized algorithms to generate non-normal correlated data (e.g., Headrick, 2002; Ruscio & Kaczetow, 2008; Vale & Maurelli, 1983) For example, imagine your goal is to generate $X$~normal, $Y$~uniform, with $\rho$=.5. Using $X_2$~uniform results in a $Y$ that is not uniform ($Y$ ends up being a linear combination of a normal and uniform). | |
| Mar 12, 2015 at 14:30 | vote | accept | Lanza | ||
| Mar 12, 2015 at 13:42 | comment | added | whuber♦ | An extensive discussion of this and related issues appears in my answer at stats.stackexchange.com/a/71303. Among other things, it makes plain that (1) the Normality assumption is irrelevant and (2) you need to make additional assumptions: the variances of $X_1$ and $X_2$ must be equal in order for the correlation of $Y$ with $X_1$ to be $\rho$. | |
| Mar 12, 2015 at 12:51 | answer | added | Artem Sobolev | timeline score: 27 | |
| Mar 12, 2015 at 12:31 | answer | added | Anthony | timeline score: 12 | |
| Mar 12, 2015 at 12:19 | history | edited | Andy | CC BY-SA 3.0 |
deleted 20 characters in body
|
| Mar 12, 2015 at 12:10 | review | First posts | |||
| Mar 12, 2015 at 12:19 | |||||
| Mar 12, 2015 at 12:06 | history | asked | Lanza | CC BY-SA 3.0 |