library(nlme) ## if(require("Hmisc")) { -- no longer: depending on ggplot2 -> total +22 pkgs is too much T.aug <- Orthodont ## now manually label(T.aug$age) <- 'anyL' ## now manually: T.aug$age <- structure(T.aug$age, label = "anyL", class = c("labelled", "numeric")) foo <- augPred(lme(distance ~ age, random = ~1|Subject, data=T.aug)) ## failed in 3.1-72 stopifnot(length(foo$age) == 1485L, inherits(foo$age, "labelled"), attr(foo$age,"label") == "anyL") ## } ## failed even if there is a variable with a class that is not being used. T.aug <- Orthodont T.aug$newage <- T.aug$age class(T.aug$newage) <- 'anyC' foo <- augPred(lme(distance ~ age, random = ~1|Subject, data=T.aug)) ## failed in 3.1-72 ## [Bug 16715] New: nlme: unable to use predict and augPredict functions in non linear mixed models ## Date: Wed, 17 Feb 2016 M1.lis <- nlsList( SSlogis, data=Soybean) ## prints "Error in qr.solve(QR.B, cc): singular matrix 'a' in solve" ## ==> MM: 2016-03-11 "fixed": now prints it as warning [you could suppress or catch] ## ==> one NA row '1989P8' in this: M1.lis ## Nonlinear Logistic Growth -- each of the 3 par. (Asym, xmid, scal) has RE term ## Model: weight ~ SSlogis(Time, Asym, xmid, scal) | Plot M1.nlme <- nlme( M1.lis ) summary(M1.nlme) ## R 3.2.2 (nlme 3.1-121) : ## Nonlinear mixed-effects model fit by maximum likelihood ## Model: weight ~ SSlogis(Time, Asym, xmid, scal) ## Data: Soybean ## AIC BIC logLik ## 1499.671 1539.881 -739.8354 ## Random effects: ## Formula: list(Asym ~ 1, xmid ~ 1, scal ~ 1) ## Level: Plot ## Structure: General positive-definite, Log-Cholesky parametrization ## StdDev Corr ## Asym 5.201186 Asym xmid ## xmid 4.197467 0.721 ## scal 1.404737 0.711 0.958 ## Residual 1.123461 ## Fixed effects: list(Asym ~ 1, xmid ~ 1, scal ~ 1) ## Value Std.Error DF t-value p-value ## Asym 19.25301 0.8031921 362 23.97062 0 ## xmid 55.01985 0.7272721 362 75.65235 0 ## scal 8.40333 0.3152893 362 26.65276 0 ## Correlation: ## Asym xmid ## xmid 0.724 ## scal 0.620 0.807 ## Standardized Within-Group Residuals: ## Min Q1 Med Q3 Max ## -6.08691772 -0.22160816 -0.03390491 0.29741145 4.84688248 ## Number of Observations: 412 ## Number of Groups: 48 M1.Fix <- fixef(M1.nlme) ## add fixed effect 'Variety' : M2.nlme <- update(M1.nlme, fixed = Asym + xmid + scal ~ Variety, start = c(M1.Fix[1], 1, M1.Fix[2], 1, M1.Fix[3], 1)) summary(M2.nlme) pred.m2 <- predict(M2.nlme, level = 0:1) stopifnot(is.data.frame(pred.m2), dim(pred.m2) == c(412, 3)) augp.m2 <- augPred(M2.nlme, level = 0:1) ## failed in nlme-3.1-124/5 stopifnot(is.data.frame(augp.m2), dim(augp.m2) == c(5308, 4) , all.equal(colMeans(augp.m2[,c("Time","weight")]), c(Time = 48.585908, weight =7.8599693), tolerance = 1e-6)# was 1e-7, 2e-9 but failed with ATLAS builds , identical(c(table(augp.m2[,".type"])), c(predict.fixed = 2448L, predict.Plot = 2448L, original = 412L)) , identical(c(table(as.vector(table(augp.m2[,".groups"])))), c("110" = 33L, "111" = 2L, "112" = 13L)) )