search for: shigesong

Dear All, I am trying to graph a simple scatter plot where the x axis is year and the y axis is a percentage (percentage of infant death). Instead of plotting the raw data, I want to plot summary statistics such as mean and median. Here is the problem: the value range of y is between 0 and 1, but since infant death is a rare event, the mean and median is very low (something like 5%), which shows

Dear All, Are there R packages that can estimate survival model with long-term survivors? This is sometimes known as "cure" model or "split-population" model. Thanks. Shige [[alternative HTML version deleted]]

Dear All, I am trying to compile&install the package "RGtk2" on my Ubuntu 10.04 box. I did not have problem with earlier versions, but with the new version, I got the following error message : ------------------------------------------------------------------------------------------------- * installing *source* package ?RGtk2? ... checking for pkg-config... /usr/bin/pkg-config

Dear All, I am trying to compile&install the package "RGtk2" on my Ubuntu 10.04 box. I did not have problem with earlier versions, but with the new version, I got the following error message : ------------------------------------------------------------------------------------------------- * installing *source* package ?RGtk2? ... checking for pkg-config... /usr/bin/pkg-config

Dear All, Is there a way to present results from multiple models in one LaTeX table? I did some google search and found out that xtable cannot automate this process (https://stat.ethz.ch/pipermail/r-help/2006-August/111174.html), are there other alternatives? What about Design and Hmisc? Thanks. Best, Shige

Dear All, Can anyone give me some hints about how to set starting values for a lmer model? For complicated models like this, good starting values can help the numerical computation and make the model converge faster. Thanks! Shige [[alternative HTML version deleted]]

Dear Colleagues, In the following simple ggplot2 code: -------------------------------- m <- ggplot(d.fig, aes(time, prob)) m + stat_summary(fun.data = "median_hilow", conf.int = .95, geom = "smooth") + facet_wrap(~ Cohort, nrow=1) + coord_cartesian(ylim = c(0, .03)) -------------------------------- Is there a way to replace the y-axis label from "prob" to

Dear All, When I tried to run the following code (taken from the *spdplyr* package vignettes): library(spdplyr) library(maptools) data(wrld_simpl) worldcorner <- wrld_simpl %>% mutate(lon = coordinates(wrld_simpl)[,1], lat = coordinates(wrld_simpl)[,2]) %>% filter(lat < -20, lon > 60) %>% dplyr::select(NAME) I got the following error messages: Error in (function (cl,

Dear All, After playing with lmer for couple of days, I have to say that I am amazed! I've been using quite some multilevel/mixed modeling packages, lme4 is a strong candidate for the overall winner, especially for multilevel generzlized linear models. Now go back to my two-level poisson model with cross-classified model. I've been testing various different model specificatios for the

Dear All, I have been using Sweave (mainly via the Sweave.sh script) and really like it. I am working a paper (using Sweave, of course) which includes several time-consuming computations, and it gets tedious to re-compile the whoel thing every time I made changes. Then I discover the "cacheSweave" package, which seems the right solution to my problem. I only have on problem. Here is

Hi all, I know this question has been asked in the past, but I am wondering if anyone running R on Linux has any guidance as to a text editor that works well with R. At the present time I am running R on Windows and using TINN-R. For a number of reasons I want to switch to Linux, but can't find much in the way of a text editor in sync with R. Any experiences, recommendations would be

Dear All, With kind help from several friends on the list, I am getting close. Now here are something interesting I just realized: for random effects, lmer reports standard deviation instead of standard error! Is there a hidden option that tells lmer to report standard error of random effects, like most other multilevel or mixed modeling software, so that we can say something like "randome

Dear All, I am doing a growth modeling using NLME. I have three levels in my data: observation, individual, household. About half of my total sample have missing values in my household-level covariates. Under this situation, the best way to go is probably to multiply impute the data (for, say, 5 times), estimate the same model separately on each model using LME function, and merge the results. My

I am running R 2.10 on SuSE 9.3. When I tried to install the package "rgl", I got error message: ... In file included from pixmap.cpp:13: pngpixmap.h: In static member function `static void PNGPixmapFormat::Load::info_callback(png_struct*, png_info*)': pngpixmap.h:149: error: invalid conversion from `long unsigned int*' to ` png_uint_32*' pngpixmap.h:149: error:

Dear All, The "coxph" function in the "survival" package allows multiple frailty terms. In all the examples I saw, however, the frailty terms are nested. What will happen if I have non-nested (that is, cross-classified) frailties in the model? Will the model still work? Do I need to take special cares when specifying these models? Thanks! Shige [[alternative HTML

Hi, I noticed that the 2.6.2 tar ball did not update, is it a good sign that no bugs have been found and it is ready to go? Shige [[alternative HTML version deleted]]

Dear All, What is the preferred way to get Bayesian analysis results (such as those from MCMCpacki, MCMCglmm, and DPpackage) into LaTeX table automatically? I have been using the "apsrtable" package and similar functions in "memisc" package, but neither seems to handle MCMC output directly. Many thanks. Shige

Dear All, Since no one has answered my previous question, let me revise it a little and ask again. My data set contains about 10,000 women born in 60 months. The outcome variable is a binary variable indicating whether one has certain health problems. My hypothesis is that the 60 months in which these women were born can be divided into three distinctive periods with respect to the binary

Dear All, I am trying to plot a lattice figure and include it in a LaTeX document via the TikZDevice package. I think the journal I am submitting to does not like colour figure, so I need to get rid of all the colours in the figure. If I directly generate PDF or EPS, the option "trellis.device(color=FALSE)", but when I do: trellis.device(color=FALSE)

Dear All, I am trying to recode a variable using the functions provided by "memisc" package. Actually I am following the examples on page 9-10 of the vignette: -------------------------------------------------------------------------- d.fig <- within(d.fig,{ sev <- recode(sev, 1 -> 0.9, 2 -> 1.0, 3 -> 1.1, 4 ->