Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Richard Goldstein <richgold@ix.netcom.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: On the joint frequency distribution |
Date | Sat, 30 Nov 2013 11:57:14 -0500 |
Syed, clearly I do not have access to your data; I have no idea what literature you are referring to I doubt the problem is due to labels have you tried graphing your data to see what is going on? Rich On 11/29/13, 4:02 PM, Syed Basher wrote: > Hi Richard, thank you for your prompt reply. The choice of 10 bins is > purely arbitrary. It could be any value between 6-10 (as seen in the > literature), or any other figure. Say, X is US stock return and Y is > Japanese stock return, by calculating the joint distribution of (i,j) > cells, we can understand the dependence structure of the two return > series, whereby numbers in bin (1,1) represent the left tail while the > numbers in bin (10,10) represent the right tail. I tried the syntax > you suggested (without doing the re-labeling), the off-diagonal > elements are zero. Is it because of not doing the re-labeling? > > Syed > > > On Fri, Nov 29, 2013 at 11:14 PM, Richard Goldstein > <richgold@ix.netcom.com> wrote: >> it is not at all clear why you want to do this, (do you have too many >> distinct values for a cross-tab? why 10 groups?, etc.) but here is, I >> think, a specific answer to your question: >> >> sort X >> egen xcat=cut(X), group(10) >> sort Y >> egen ycat=cut(Y), group(10) >> ta xcat ycat >> >> you might want to re-label the new variables >> >> Rich >> >> On 11/29/13, 2:45 PM, Syed Basher wrote: >>> Dear Stata Users, >>> >>> I am using Stata 12.1. I have two series X and Y (2280 obs each). I >>> want to compute the joint frequency distribution of the two variables >>> (i.e., cross tabulation). First, I want to sort the variables in >>> ascending order. Second, I would like to divide each series evenly >>> into 10 bins. Finally, I want to get the joint frequency distribution >>> of the two series for all combination of (i,j) cells (that is, the >>> diagonal and off-diagonal elements). Problem is how can I do this in >>> Stata. Your help, hint or suggestion is highly appreciated. >>> >>> Regards, >>> Syed Basher >>> Doha, Qatar * * 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/