Re: [R] Time Series on Binary Data.
Among various possibilities, you might consider a logistic or probit regression model with ARMA errors specified via Gaussian copula. This approach is implemented in the package gcmr (Gaussian Copula Marginal Regression). Example: logistic model with covariates S1 and S2 and AR(1,2) errors fit - gcmr( Response ~ S1 + S2 + ... , data = your.data, marginal = binomial.marg, cormat = arma.cormat(1,2), options( nrep=1000 ) ) See help(gcmr) for more details. Cristiano - Cristiano Varin cristiano.va...@unive.it Department of Environmental Sciences, Informatics and Statistics Ca' Foscari University of Venice San Giobbe, Cannaregio 873, 30121 Venezia, Italy Tel: +39 0412347439 Fax: +39 0412347444 http://cristianovarin.weebly.com Hi, I have a dichotomous data where some my independent variables are categorical, some are continuous and some are binary(0/1) My dependent is a binary response (Fail/NoFail,0/1) . The data is some readings collected everyday over a period of time. The goal is to use this data and see if we can figure out cause of failure ,the end response. Example data format Date, Type,Mileage,S1,S2,S3 , Response 03/02/2013,A,32000,1,0,1,.., 1 03/03/2013,B,32400,0,0,0,...,0 03/04/2013,C,45000,0,1,1,..,1 Can we do Time series modeling?? Any suggesstions on what type of other exploratory analysis can be used to figure out patterns in data ?? Thanks shi __ R-help@r-project.org mailing list 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.
Re: [R] Fit Negbin glm model with autoregressive correlation structure
You may also introduce ARMA errors in negative binomial regression via Gaussian copulas. The following webpage illustrates how to use the package gcmr (Gaussian Copula Marginal Regression) for fitting negative binomial regression model with ARMA(2,1) errors: http://cristianovarin.weebly.com/gcmr.html With best regards, Cristiano Varin - Cristiano Varin cristiano.va...@unive.it Department of Environmental Sciences, Informatics and Statistics Ca' Foscari University of Venice http://cristianovarin.weebly.com Il 31/01/2014 09:41, flavie.v...@vetsuisse.unibe.ch ha scritto: Hello, I am attempting to estimate the effect of various variables on the time-series of counts of reported cattle stillbirths. We investigate the effect of day-of-week, month, holidays etc...and also the effect of non-temporal variables. We performed model comparisons between Gaussian glm, Poisson glm and negbin glm and the latter seems most appropriate for our data. We found that the residuals from our best model are not i.i.d. but follow an autoregressive process of order 5 , AR (5). I therefore wish to re-run this model after adding an AR(5) correlation structure in order to get unbiased estimates and standard errors for the variables retained in the model. In the past, I have been faced with a similar situation for a Gaussian glm and used the gls function with a corStruct object describing the within-group correlation structure. However, this would not work with our negbin model. Looking around on various help forums, I came across the possibilities of using generalized estimating equations instead. The gee function (in gee package) has a corstr object which would allow me to specify an AR process of whichever order but there is no option to include a negbin family. The geeglm function (in geepack package) does recognized the negbin family but only gives an option to fit an AR(1) correlation structure. The corstr object seems to have a userdefined option but it is unclear how it could be defined for an AR(5) process. In short, my questions are: *is it possible to include an AR (p) correlation structure directly into a negbin glm? How? * if GEE are the way forward, how can the the corstr object in geeglm be defined for an AR(p) process? Thank you for your suggestions, Dr Flavie Vial Veterinary Public Health Institute DCR-VPH, Vetsuisse Fakultät Schwarzenburgstrasse 155 CH-3003 Bern Switzerland flavie.v...@vetsuisse.unibe.chmailto:flavie.v...@vetsuisse.unibe.ch [[alternative HTML version deleted]] __ R-help@r-project.org mailing list 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 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] Output of silhouette (cluster package)
Dear R users, I am mailing you about the graphical output of silhouette (cluster package) From the example of silhouette in help(silhouette): ar - agnes(ruspini) si3 - silhouette(cutree(ar, k = 5), # k = 4 gave the same as pam() above +daisy(ruspini)) plot(si3, nmax = 80, cex.names = 0.5) from which one may conclude that group 1 is composed by units from 1 to 20, group 2 by units from 21 to 43, group 3 by units from 44 to 57, group 4 by units from 58 to 60 and, finally, group 5 by units from 61 to 75. However, this seems to be in contrast with the output of silhouette where the fourth group is composed by units from 46 to 48 instead of units from 58 to 60 (belonging to the third cluster), see si3 cluster neighbor sil_width [1,] 15 0.679838078 [2,] 15 0.745615002 [3,] 15 0.758796123 [4,] 14 0.715554768 [5,] 15 0.664657114 [6,] 14 0.783993831 [7,] 12 0.590057470 [8,] 14 0.747969458 [9,] 15 0.792304760 [10,] 14 0.803547635 [11,] 14 0.742402051 [12,] 14 0.722302731 [13,] 14 0.665412622 [14,] 15 0.756910666 [15,] 15 0.700685403 [16,] 15 0.743601834 [17,] 15 0.614854124 [18,] 15 0.708007860 [19,] 15 0.700093839 [20,] 14 0.568989067 [21,] 24 0.751866935 [22,] 24 0.790783667 [23,] 24 0.802659788 [24,] 24 0.785895823 [25,] 24 0.822943473 [26,] 24 0.831313347 [27,] 24 0.818043337 [28,] 24 0.805454305 [29,] 24 0.770547118 [30,] 24 0.768289979 [31,] 23 0.794485567 [32,] 24 0.829925955 [33,] 24 0.807379640 [34,] 24 0.790626589 [35,] 24 0.817427927 [36,] 23 0.793572412 [37,] 24 0.760561408 [38,] 24 0.743170109 [39,] 23 0.761413953 [40,] 23 0.704193051 [41,] 24 0.297007126 [42,] 24 0.522049838 [43,] 23 0.488556828 [44,] 34 0.377632488 [45,] 34 0.007214464 [46,] 43 0.699407534 [47,] 43 0.837451212 [48,] 43 0.794349431 [49,] 34 0.632862996 [50,] 34 0.586149139 [51,] 34 0.647326133 [52,] 34 0.650020368 [53,] 34 0.629131005 [54,] 34 0.618843633 [55,] 34 0.586439350 [56,] 34 0.586788051 [57,] 34 0.668108812 [58,] 34 0.650074540 [59,] 34 0.628444500 [60,] 34 0.591393005 [61,] 51 0.770110294 [62,] 51 0.815309198 [63,] 54 0.771622667 [64,] 51 0.806125429 [65,] 51 0.850310507 [66,] 51 0.822984066 [67,] 51 0.852743923 [68,] 51 0.762055943 [69,] 51 0.839180986 [70,] 51 0.854894699 [71,] 51 0.838106473 [72,] 51 0.774812117 [73,] 51 0.795021304 [74,] 51 0.759681469 [75,] 51 0.742553847 attr(,Ordered) [1] FALSE attr(,call) silhouette.default(x = cutree(ar, k = 5), dist = daisy(ruspini)) attr(,class) [1] silhouette Thanks for your attention, Cristiano - Cristiano Varin [EMAIL PROTECTED] http://www.dst.unive.it/~sammy/ [[alternative HTML version deleted]] __ R-help@r-project.org mailing list 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.
