[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 |
st: RE: RE: percentage for each month plotted vs. month |

Date |
Mon, 2 May 2005 15:45:20 +0100 |

In practice, this would lead to lots of overplotting. Leaner graph files would ensue from gen month = month(date) egen adult_x = mean(100 * (2 - x)) if adult == 1, by(month) egen child_x = mean(100 * (2 - x)) if adult == 2, by(month) bysort month (adult_x) : replace adult_x = adult_x[1] if mi(adult_x) bysort month (child_x) : replace child_x = child_x[1] if mi(child_x) egen tag = tag(month) line *_x month if tag Nick n.j.cox@durham.ac.uk Nick Cox > I assume variables > > date: Stata daily date > x: as defined below > adult: as defined below > > You can get months by > > gen month = month(date) > > The percent of drug use would then be > > egen adult_x = mean(100 * (2 - x)) if adult == 1, by(month) > > egen child_x = mean(100 * (2 - x)) if adult == 2, by(month) > > The graph would be > > line *_x month > > To cycle over many months, > > foreach v of var x y z a b c d e f { > egen adult_`v' = mean(100 * (2 - `v')) if adult == 1, by(month) > egen child_`v = mean(100 * (2 - `v')) if adult == 2, by(month) > line *_`v' month, saving(`v') > more > } > > Note how binary coding 1 for use, 0 for no use, would lead > to cleaner formulae. > > Nick > n.j.cox@durham.ac.uk > > Richard_Lenhardt@rush.edu > > > I have the following survey data of 1000 patients collected over 15 > > months: > > > > Each observation contains the following variables: > > X: "1" for using drug x and "2" for not using drug x > > DATE: Day and month patient was enrolled in study (variable > > DATE as day > > month year) > > Adult: "1" yes "2" no > > > > I would like to plot the percentage of patients who were > > using drug x in a > > given month vs. month & year. So the axes would be Y: "percent of > > patients using drug x in given month" and X axis: "month & year of > > enrollment" > > > > I cannot seem to figure out an easy way to do this. It seems > > that I have > > to generate many variables to get a simple percentage of > > patients who were > > using drug x for that particular month and then to plot it. > > > > Also, I would like to get more sophisticated as well: the > > same graph as > > above but have two lines on one graph; one line would > > represent the data > > for children and the other for adults. > > > > Also, I'd like to generate many graphs using the do editor > > (same as above > > for other drugs). When I try using the do editor this, I get > > one graph > > and then the next is shown and I get to save only the last. > > Is there a > > way I could run several graphs using the do editor and then > > save them all > > so that I could view them leisurely? > > * * 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: RE: tabl question (table of reg results)** - Next by Date:
**st: re: discerning a variable's "type"** - Previous by thread:
**st: RE: tabl question (table of reg results)** - Next by thread:
**st: ivreg with clustering** - Index(es):

© Copyright 1996–2016 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |