Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
Stas Kolenikov <[email protected]> |

To |
"[email protected]" <[email protected]> |

Subject |
st: visualize -svy : tab x y- |

Date |
Wed, 15 Jan 2014 13:10:43 -0600 |

Folks, Suppose I want to visualize something like webuse nhanes2, clear svy : tab race hlthstat, se row in a way similar to what -marginsplot- does. The first natural idea that I thought of was to run svy: proportion hlthstat, over(race) only to discover that -margins- does not support -proportion- (although this seems like a no-brainer for -margins- to work on -- just redisplay verbatim what -proportion- did). My next step was to try svy : mlogit hlthstat i.race which essentially is a weird way to compute the log odds ratios against a given category. Now, -margins- does work, but it only gives the proportions for a single outcome, like hlthstat==1, and I cannot seem to find a way to produce several margins over several equations (although that makes perfect sense for -mlogit-), and hence plot them with -marginsplot- afterwards. In -marginsplot- terms, this could be something like a dimension to plot over, so may be another possible -dimlist- could be equations in a multi-equation command. (Consider this a feature request!) I am sure the question of how to visualize the cross-tabs is not that far-fetched, so there should be something Phil Enders or Michael Mitchell have produced... that I am missing here. There was a discussion as far back as 2006 (http://www.stata.com/statalist/archive/2006-09/msg00506.html), but -margins- is a much later development, as far as I can recall, so the ways to adapt -margins- may have been discovered since then. -- Stas Kolenikov, PhD, PStat (ASA, SSC) -- Senior Survey Statistician, Abt SRBI -- Opinions stated in this email are mine only, and do not reflect the position of my employer -- http://stas.kolenikov.name * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/

- Prev by Date:
**Re: st: Cox model with a segmented time-dependent covariate** - Next by Date:
**Re: st: Cox model with a segmented time-dependent covariate** - Previous by thread:
**st: Seemingly unrelated regressions and dynamic panel data methods** - Next by thread:
**st: Counting firms in a panel dataset** - Index(es):