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

From |
"Nick Cox" <n.j.cox@durham.ac.uk> |

To |
<statalist@hsphsun2.harvard.edu> |

Subject |
RE: st: create dummy variables for every 5 values |

Date |
Fri, 11 Nov 2005 10:39:22 -0000 |

Your last line was gen dummy = ceil(_n/5) and I misunderstood that to be the end of your solution. With panel data, and the possibility of unbalanced panels, gaps in data, etc. a more general solution would be to work on the year variable, not _n. gen dummy = ceil(year/5) tab dummy, gen(dummy) satisfies the original example. Nick n.j.cox@durham.ac.uk Sergio Correia > Well, the idea is to make the dummies from that variable using the > tab dummy, g(dummy) > command that Lynda described. The recode alternative is useful except > if you have many years and therefore require many dummies, for each > 5-year period. Nick Cox > > I doubt that this is what is wanted here. > > Typically, unless the dataset contains just > > 10 observations, it will give a dummy variable > > that is not a dummy variable. Even in Sergio's > > toy example it takes on 6 distinct values. > > > > Lynda mentioned a solution, which can be tweaked to > > > > gen yr=year > > recode yr (1991/1995=0)(1996/2000=1) > > > > The -tabulate- does nothing useful as -yr- is already a dummy. > > > > In this situation, I would not use -recode-, but > > that's mostly taste. Other solutions are > > > > gen dummy = 0 if inrange(year,1991,1995) > > replace dummy = 1 if inrange(year,1996,2000) > > > > or > > > > gen dummy = cond(inrange(year,1991,1995), 0, > cond(inrange(year,1996,2000), 1,. ) ) > > > > but there's not much to choose between solutions, so long > > as you don't falsely map any years not 1991-2000 to either 0 or 1. Sergio Correia > > > Try this: > > > > > > 1) create example file > > > set obs 30 > > > gen year=1990+_n > > > > > > 2) answer > > > gen dummy = ceil(_n/5) > > > > > > The trick is in using the ceil function with the _n variable. Lynda Zhang > > > > I have a variable Year: > > > > ... > > > > 1991 > > > > 1992 > > > > 1993 > > > > 1994 > > > > 1995 > > > > 1996 > > > > 1997 > > > > 1998 > > > > 1999 > > > > 2000 > > > > ... > > > > > > > > How do I create one dummy variable for each 5-year > interval with one > > > > command. > > > > > > > > Typically I can get it manually by generating one dummy > at a time > > > > gen year1=1 if year>=1991 & year<=1995 > > > > ... > > > > > > > > or using Recode: > > > > gen yr=year > > > > recode yr (1991/1995=1)(1996/2000=2) > > > > tab yr, g(y) * * 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/

- Prev by Date:
**st: Interpreting cumulative sum output** - Next by Date:
**RE: st: Adding lines to graphs** - Previous by thread:
**Re: st: create dummy variables for every 5 values** - Next by thread:
**st: missing value imputation in Fixed Effects GLS model** - Index(es):

© Copyright 1996–2014 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |