- 1. Forecasting ARIMA(1,1,0) (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Thu, 3 May 2001 13:46:24 +0200
- Dear all, Sorry for a long explanation but I really got lost in trying to forecast a non-stationary ARIMA(1,1,0) series. My problem is - why does the forecast not follow the trend? x <- c(2.59,2.57,2
- /archives/html/s-news/2001-05/msg00031.html (8,399 bytes)
- 2. initial values in nls (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Fri, 19 Jan 2001 13:46:42 +0100
- Dear all, When starting nonlinear regression, one has to indicate the initial values of the coefficients. How can I do this in FFF <- data.frame(x, y) param(FFF, "a1") <- fparam[1] param(FFF, "a1") <
- /archives/html/s-news/2001-01/msg00135.html (6,874 bytes)
- 3. initial values in nls (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Tue, 23 Jan 2001 17:08:52 +0100
- Here is my (slightly edited) original problem and responses from Douglas Bates and Nick Ellis (I thank both of them): When starting nonlinear regression, one has to indicate the initial values of the
- /archives/html/s-news/2001-01/msg00152.html (8,479 bytes)
- 4. [S] plot.gam (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Wed, 8 Dec 1999 21:11:58 +0100
- Dear all, How can one explain why the following two graphs differ? The first graph: The second graph: Thank you. Remigijus -- This message was distributed by s-news@wubios.wustl.edu. To unsubscribe s
- /archives/html/s-news/1999-12/msg00057.html (7,540 bytes)
- 5. [S] plot.gam (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Thu, 9 Dec 1999 09:05:44 +0100
- Many thanks to all who responded to my question Basically, the answer was - masking. Many proposed to use find(ethanol), find("NOx"), find("E"), find("C). Actually, rm(ethanol,NOx,E,C) helped. Remigi
- /archives/html/s-news/1999-12/msg00062.html (7,810 bytes)
- 6. [S] accuracy of binomial... (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Mon, 25 Oct 1999 20:04:21 +0159
- Dear S+ers, In order to find how accurately S+ calculates binomial probabilities I wrote the following program (for S-Plus for Windows, Version 3.1 Release 3): function() { dens <- numeric(18) half <
- /archives/html/s-news/1999-10/msg00203.html (8,618 bytes)
- 7. [S] accuracy of binomial... (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Wed, 27 Oct 1999 18:03:39 +0200
- Many thanks to all who responded to my question about the accuracy of statistical distributions in S-Plus. The most useful (to me) answer was by Bruce McCullough who sent me to www.stat.uni-muenchen.
- /archives/html/s-news/1999-10/msg00222.html (7,560 bytes)
- 8. msg00188.html (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Wed, 21 Apr 1999 13:53:50 +0200
- Hello s-news, I believe that if the term poly(svoris,atst,2)0.2 is dropped, the updated model will be better (say, in Cp sense). How to do this? Call: lm(formula = kaina ~ poly(svoris, atst, 2), data
- /archives/html/s-news/1999-04/msg00188.html (7,662 bytes)
- 9. msg00218.html (score: 1)
- Author: Remigijus Lapinskas <remigijus.lapinskas@maf.vu.lt>
- Date: Fri, 23 Apr 1999 23:39:17 +0200
- Hello All, I want to export an S-Plus character data set to the *.dbf file. However, it seems that the character fields in the resulting *.dbf file will always have the same length of 21. What should
- /archives/html/s-news/1999-04/msg00218.html (7,280 bytes)
- 10. [S] integrate (score: 1)
- Author: "Remigijus Lapinskas" <Remigijus.Lapinskas@maf.vu.lt>
- Date: Fri, 14 Aug 1998 16:57:32 +0000
- Hello..I am using S-PLUS For Windows Version 3.1 Release 3. What's wrong with the following program: j_4 ff_function(x,j){sin(x*j)} integrate(ff,lower=0,upper=1,j) or, in other words, how to pass the
- /archives/html/s-news/1998-08/msg00110.html (7,005 bytes)
This search system is powered by
Namazu