[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
UENOHARA Hideaki <uhideaki@td6.so-net.ne.jp> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: st: legend of -graph pie- |

Date |
Sat, 01 Nov 2003 15:30:11 +0900 |

Dear Vince and all. I appreciate Stata Corp.'s prompt action, but I'm afraid the peculiar behavior of -graph pie- still remains. It still includes unintended keys in the legend. Of course I recognized that the latest executable and ado-files (31 Oct 2003) are installed in my computer. With the following commands, can anybody replicate this behavior? . sysuse auto . graph pie, over(rep78) by(foreign) H.U. On Fri, 31 Oct 2003 10:37:16 -0600 vwiggins@stata.com (Vince Wiggins, StataCorp) wrote: > UENOHARA Hideaki <uhideaki@td6.so-net.ne.jp> notes that -graph pie- includes a > key in the legend for missing values when the -over()- and -by()- options are > combined and the over group variable has missing values > > > After updating the executable and ado-files, I have found -graph > > pie- behaves peculiarly with respect to the legend. > > > > The problem is as follows: > > > > . sysuse auto.dta > > . graph pie, over(rep78) by(foreign) > > > > produces a pie chart whose legend has an extra key which corresponds > > to missing values. On the other hand > > > > . graph pie, over(rep78) > > > > does not. > > Friedrich Huebler <huebler@rocketmail.com> notes that this behavior extends to > including keys for groups excluded with an -if- condition and also suggests > that we can use -legend(order(...))- to exclude any keys we wish. > > This behavior is an unintended consequence of the fact that -by()- tries to be > completely generic and have no knowledge of the -graph- subcommand being > -by-ed, while at the same time allowing a common legend with a single set of > keys to be produced. Suffice to say that we cannot get rid of those extra > keys without making -by()- smarter about which graph subcommand is being run. > That intelligence will be added in the next official update, either late this > afternoon or Monday. Keys for unwanted missings and explicitly excluded > groups will no longer haunt our legends. > > > -- Vince > vwiggins@stata.com > > * > * For searches and help try: > * http://www.stata.com/support/faqs/res/findit.html > * http://www.stata.com/support/statalist/faq > * http://www.ats.ucla.edu/stat/stata/ > Yours Sincerely, UENOHARA Hideaki uhideaki@td6.so-net.ne.jp * * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**Follow-Ups**:**st: Re: legend of -graph pie-***From:*Friedrich Huebler <huebler@rocketmail.com>

- Next by Date:
**Re: st: programming question** - Next by thread:
**st: Re: legend of -graph pie-** - Index(es):

© Copyright 1996–2015 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |