> On 12 May 2017, at 16:40 , Tobias Christoph <s3toc...@uni-bayreuth.de> wrote: > > Hey guys, > > thanks a lot for your tips. The regression is finally running. As you > said, I had to integrate the column "year" in the function "time" in R. > > So I used the following formula: *plm(log(revenue) ~ log(supply) + > factor(town)*time(year), data=R_Test_log_Neu)*
Um, that might not do what I think you think it does. time() gives you the "vector of times at which a time series was sampled". If you feed it any regular vector, it just gives the numbers 1:n, witness > time(rnorm(20)) [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 attr(,"tsp") [1] 1 20 1 I suspect you just want "year" in the formula. > > So I have now sucessfully added a linear trend to my regression model? > Another question that concernes me is how to add a quadratic trend > instead of a linear trend. Can I just square the column "year"? In principle, yes, but as others have pointed out, centering the variable may be a good idea, for numerical stability. -pd > > Enjoy your weekend, > > Toby > > _My results see below:_ > > Balanced Panel: n=11, T=12, N=132 > > Residuals : > Min. 1st Qu. Median 3rd Qu. Max. > -0.09610 -0.02370 -0.00152 0.01980 0.14000 > > Coefficients : > Estimate Std. Error t-value Pr(>|t|) > log(supply) -0.0080702 0.0133675 -0.6037 0.547411 > factor(town)2:time(year) -0.0063245 0.0053744 -1.1768 0.242100 > factor(town)3:time(year) 0.0295522 0.0056776 5.2050 1.053e-06 *** > factor(town)4:time(year) 0.0062191 0.0054152 1.1485 0.253549 > factor(town)5:time(year) 0.0159028 0.0054954 2.8939 0.004681 ** > factor(town)6:time(year) 0.0237112 0.0055395 4.2804 4.316e-05 *** > factor(town)7:time(year) 0.0410007 0.0055734 7.3565 5.576e-11 *** > factor(town)8:time(year) 0.0239085 0.0053751 4.4480 2.271e-05 *** > factor(town)9:time(year) 0.0242342 0.0056855 4.2625 4.619e-05 *** > factor(town)10:time(year) 0.0105890 0.0053302 1.9866 0.049733 * > factor(town)11:time(year) 0.0095270 0.0056354 1.6906 0.094065 . > --- > Signif. codes: 0 �***� 0.001 �**� 0.01 �*� 0.05 �.� 0.1 � � 1 > > Total Sum of Squares: 0.46388 > Residual Sum of Squares: 0.2001 > R-Squared: 0.56863 > Adj. R-Squared: 0.4292 > F-statistic: 11.8637 on 11 and 99 DF, p-value: 7.3065e-14 > > > > Am 11.05.2017 um 23:17 schrieb David L Carlson: >> What Rui said, but as important, you have four columns in your data called >> "town", "year", "revenue", and "supply". You do not have a column called >> "time". >> >> ------------------------------------- >> David L Carlson >> Department of Anthropology >> Texas A&M University >> College Station, TX 77840-4352 >> >> -----Original Message----- >> From: R-help [mailto:r-help-boun...@r-project.org] On Behalf Of Rui Barradas >> Sent: Thursday, May 11, 2017 3:36 PM >> To: Tobias Christoph <s3toc...@uni-bayreuth.de>; Duncan Murdoch >> <murdoch.dun...@gmail.com>; r-help@r-project.org >> Subject: Re: [R] Error: invalid type (closure) for the variable 'time' - >> object specific trend >> >> Hello, >> >> A closure is, like you say, a function. >> At an R prompt try: >> >>> typeof(time) >> [1] "closure" >> >> So like Duncan suggested rename 'time', for instance capitalize it >> 'Time'. That should do it. >> >> Hope this helps, >> >> Rui Barradas >> >> >> >> Em 11-05-2017 21:20, Tobias Christoph escreveu: >>> Hey Duncan, >>> >>> thank you very much for your quick reply. >>> >>> _My data used:_ >>> >>> 1st column(town):1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,...........,11 >>> >>> 2nd column(year):1,2,3,4,5,6,7,8,9,10,11,12,1,2,3.......,12 >>> >>> 3rd column (revenue): >>> >>> 4th colum (supply): >>> >>> I have now renamed my colums and did the regression again. Now there is >>> a problem with R-squared, as it is the sum of 1 now with no given std. >>> error and t-value. This is probably due to the fact, that I try to >>> estimate more parameters than data. >>> >>> To add a linear trend I found the following formular:*lm(y ~ x1 + >>> factor(ccode)*time, data=df)* >>> >>> I try to I decode it for and use it for my regression: *plm(log(revenue) >>> ~ log(supply) + factor(town)*time, data=R_Test_log_Neu)* >>> >>> When I do this regression I will get the original error: "invalid type >>> (closure) for the variable 'time' - object specific trend" >>> >>> With the notation"time" not my colum is meant, but probably the command >>> "time" in R. >>> >>> Can you follow my thoughts? >>> >>> Tobi >>> >>> >>> >>> >>> >>> >>> Am 11.05.2017 um 17:23 schrieb Duncan Murdoch: >>>> Duncan Murdoch >>> ______________________________________________ >>> R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see >>> https://stat.ethz.ch/mailman/listinfo/r-help >>> PLEASE do read the posting guide >>> http://www.R-project.org/posting-guide.html >>> and provide commented, minimal, self-contained, reproducible code. >> ______________________________________________ >> R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see >> https://stat.ethz.ch/mailman/listinfo/r-help >> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html >> and provide commented, minimal, self-contained, reproducible code. >> > > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. -- Peter Dalgaard, Professor, Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Office: A 4.23 Email: pd....@cbs.dk Priv: pda...@gmail.com ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.