Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Nick Cox <n.j.cox@durham.ac.uk> |
To | "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |
Subject | st: RE: Lag Operator |
Date | Fri, 24 Jun 2011 10:31:06 +0100 |
Suppose I observe firm Raychaudhuri Corporation at times 1, 2, 3 and 5, so that there is a gap in the data at time 4. The gap is what bites for this and other examples. _n-1 is the subscript for the previous observation. After -sort-ing appropriately, the previous observation for time 5 is that for time 3. L. gives you the value for the previous time. For time 5, that would be time 4, but time 4 is not present in the data, so L. for any variable will return missing for time 5. Nick n.j.cox@durham.ac.uk Jibonayan Raychaudhuri I am a bit confused about the way the L (lag) operator behaves in Stata. For a panel of firms I first create the following lagged varaibles: #delimit; iis compname; tis newyear; gsort +compname +newyear; #delimit; local varlist1 sales forexsales; foreach var of local varlist1{; by compname: generate L1log`var'= log`var'[_n-1]; by compname: generate L2log`var'= log`var'[_n-2]; by compname: gen D1log`var'= log`var'[_n]-log`var'[_n-1]; }; Next I create : #delimit; iis compname; tis newyear; gsort +compname +newyear; xtset compname newyear; #delimit; local varlist1 sales forexsales; foreach var of local varlist1{; by compname: generate L1log`var'_n= L1.log`var'; by compname: generate L2log`var'_n= L2.log`var'; by compname: gen D1log`var'_n= D1.log`var'; }; The two do not give me the same result. I find a something more confusing on listing the variables (please see row 8) : Any help appreciated. Best, Jibonayan list compname nic newyear logforexsales L1logforexsales L1logforexsales_new in 1/30 +-----------------------------------------------------------------------------------------+ | compname nic newyear logfo~les L1~xsales L1logfo.. | |-----------------------------------------------------------------------------------------| 1. | 20 Microns Ltd. 269 1997 -1.752332 . . | 2. | 20 Microns Ltd. 269 1998 -1.930396 -1.752332 -1.752332 | 3. | 20 Microns Ltd. 269 1999 -2.167871 -1.930396 -1.930396 | 4. | 20 Microns Ltd. 269 2000 -2.09666 -2.167871 -2.167871 | 5. | 20 Microns Ltd. 269 2001 -2.181683 -2.09666 -2.09666 | |-----------------------------------------------------------------------------------------| 6. | 20 Microns Ltd. 269 2002 -1.852771 -2.181683 -2.181683 | 7. | 20 Microns Ltd. 269 2003 -1.910348 -1.852771 -1.852771 | 8. | 20 Microns Ltd. 269 2005 -1.855497 -1.910348 . | 9. | 20 Microns Ltd. 269 2006 -2.300922 -1.855497 -1.855497 | 10. | A B B Instrumentation Ltd. [Merged] 331 1990 -3.08322 . . | |-----------------------------------------------------------------------------------------| 11. | A B B Instrumentation Ltd. [Merged] 331 1991 -3.543726 -3.08322 -3.08322 | 12. | A B B Instrumentation Ltd. [Merged] 331 1992 -2.775719 -3.543726 -3.543726 | 13. | A B B Instrumentation Ltd. [Merged] 331 1993 -3.566005 -2.775719 -2.775719 | 14. | A B B Instrumentation Ltd. [Merged] 331 1994 -3.640079 -3.566005 -3.566005 | 15. | A B B Instrumentation Ltd. [Merged] 331 1995 -2.996687 -3.640079 -3.640079 | |-----------------------------------------------------------------------------------------| 16. | A B B Instrumentation Ltd. [Merged] 331 1996 -2.698229 -2.996687 -2.996687 | * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/