similar to: survey 3.20

Version 3.18 of the survey package is now percolating through CRAN. Since the last announcement on this list, in February, the main additions are - standard errors for survival curves (both Kaplan-Meier and Cox model) - svyciprop() for confidence intervals on proportions, especially in small samples or near 0 or 1. - predictive margins by direct standardization, with marginpred() -

Version 3.18 of the survey package is now percolating through CRAN. Since the last announcement on this list, in February, the main additions are - standard errors for survival curves (both Kaplan-Meier and Cox model) - svyciprop() for confidence intervals on proportions, especially in small samples or near 0 or 1. - predictive margins by direct standardization, with marginpred() -

Version 3.3 of "survey" is percolating through CRAN. Since the last announcement on this list, version 2.9, the main additions are calibration estimators: linear, bounded linear, raking ratio, bounded raking ratio, logit. Other updates and bug fixes are described in http://faculty.washington.edu/tlumley/survey/NEWS -thomas Thomas Lumley Assoc. Professor,

Version 3.3 of "survey" is percolating through CRAN. Since the last announcement on this list, version 2.9, the main additions are calibration estimators: linear, bounded linear, raking ratio, bounded raking ratio, logit. Other updates and bug fixes are described in http://faculty.washington.edu/tlumley/survey/NEWS -thomas Thomas Lumley Assoc. Professor,

Version 1.9 of the survey package, now percolating through CRAN, adds a beta implementation of replication weights. These can either be created from a survey design (using BRR, JK1, or JKn schemes) or provided by the user. These have been tested on only a few examples so far: there seem to be relatively few published datasets with suitable analyses. As with earlier versions of the package, I

Version 1.9 of the survey package, now percolating through CRAN, adds a beta implementation of replication weights. These can either be created from a survey design (using BRR, JK1, or JKn schemes) or provided by the user. These have been tested on only a few examples so far: there seem to be relatively few published datasets with suitable analyses. As with earlier versions of the package, I

'surveyNG' version 0.3 is on CRAN. This package provides experimental features for survey analysis that may be incorporated in the survey package in the future. Currently there are facilities for analysis of complex surveys using (possibly large) data sets stored in a SQLite database. However, analysis facilities for these SQL-backed survey designs are rather more limited than in the

'surveyNG' version 0.3 is on CRAN. This package provides experimental features for survey analysis that may be incorporated in the survey package in the future. Currently there are facilities for analysis of complex surveys using (possibly large) data sets stored in a SQLite database. However, analysis facilities for these SQL-backed survey designs are rather more limited than in the

Version 3.11 of the survey package is making its way through CRAN. Since the last announcement on this list, of version 3.9, last September, there have been many minor bug fixes and usability improvements. The main new features are - loglinear models with svyloglin() - database-backed designs now allow new variables to be created, support ODBC in addition to DBI database interfaces, and

Version 3.11 of the survey package is making its way through CRAN. Since the last announcement on this list, of version 3.9, last September, there have been many minor bug fixes and usability improvements. The main new features are - loglinear models with svyloglin() - database-backed designs now allow new variables to be created, support ODBC in addition to DBI database interfaces, and

Version 2.9 of survey is on CRAN. In addition to various minor improvements and bug fixes there are two major changes - full multistage finite-population sampling is supported (as in SUDAAN) - the same analysis commands can be used for all design types (eg svymean instead of svrepmean for replicate weight designs) -thomas Thomas Lumley Assoc. Professor, Biostatistics tlumley at

Version 2.9 of survey is on CRAN. In addition to various minor improvements and bug fixes there are two major changes - full multistage finite-population sampling is supported (as in SUDAAN) - the same analysis commands can be used for all design types (eg svymean instead of svrepmean for replicate weight designs) -thomas Thomas Lumley Assoc. Professor, Biostatistics tlumley at

-------------------------------------------------- From: "Ahoussou Sylvie" <sylvie.ahoussou at antilles.inra.fr> Sent: Friday, September 12, 2008 9:48 AM To: "Thomas Lumley" <tlumley at u.washington.edu> Subject: Re: [R] Complex sampling survey _ Use of survey package > Thanks for your answer > > I think I made a mistake when I recopied the 5 first rows of

Version 3.9 of the survey package is now on CRAN. Since the last announcement (version 3.6-11, about a year ago) the main changes are - Database-backed survey objects: the data can live in a SQLite (or other DBI-compatible) database and be loaded as needed. - Ordinal logistic regression - Support for the 'mitools' package and multiply-imputed data - Conditioning plots,

Version 3.9 of the survey package is now on CRAN. Since the last announcement (version 3.6-11, about a year ago) the main changes are - Database-backed survey objects: the data can live in a SQLite (or other DBI-compatible) database and be loaded as needed. - Ordinal logistic regression - Support for the 'mitools' package and multiply-imputed data - Conditioning plots,

Hi, All Thanks to Robert Baskin, Thomas Lumley, and Spencer Graves for the valuable helps. I have learned a lot from this discussion. I put all discussions together without editing, so we can see how things are evolved. Likely, I have a lot of articles to read. As in the discussion, mixed modeling approach is a poosible but may be over-kill in my posted data analyses. I will explore other

New versions of leaps and biglm are percolating through CRAN. The new version of biglm fixes a bug in sandwich standard errors with weights, and adds predict(), deviance() and AIC() methods [based on code from Christophe Dutang]. The new version of leaps adds a regsubsets() method for biglm objects, so that the subset selection algorithms can be run efficiently on large data sets. -thomas

New versions of leaps and biglm are percolating through CRAN. The new version of biglm fixes a bug in sandwich standard errors with weights, and adds predict(), deviance() and AIC() methods [based on code from Christophe Dutang]. The new version of leaps adds a regsubsets() method for biglm objects, so that the subset selection algorithms can be run efficiently on large data sets. -thomas

Version 3.26 of the survey package is percolating through CRAN. Since the last announcement on this list, of version 3.20, about 18 months ago, the main changes are -- an option to calculate replicate-weight variances from sums of squares around the point estimate rather than from the variance of the replicates ("MSE" style) -- Preston's multistage rescaled bootstrap,

Dear Thomas, Thank you, yes, that sounds good, and I take the point about integer overflow. Various questions: (a) Is there some way I can try out the routine with this modification? (I am on a Linux system where I am just a user - I cannot install new versions of software myself) ? (b) Is there a reference you can give me to a published paper where the method being used to compute the