Description Usage Arguments Value References See Also Examples

Multiple outputation is a procedure whereby excess observations are repeatedly randomly sampled and discarded. The method was originally developed to handle clustered data where cluster size is informative, for example when studying pups in a litter. In this case, analysis that ignores cluster size results in larger litters being over-represented in a marginal analysis. Multiple outputation circumvents this problem by randomly selecting one observation per cluster. Multiple outputation has been further adapted to handle longitudinal data subject to irregular observation; here the probability of being retained on any given outputation is inversely proportional to the visit intensity. This function creates a single outputted dataset.

1 | ```
outputation(data, weights, singleobs, id, time, keep.first)
``` |

`data` |
the original dataset on which multiple outputation is to be performed |

`weights` |
the weights to be used in the outputation, i.e. the inverse of the probability that a given observation will be selected in creating an outputted dataset. Ignored if singleobs=TRUE |

`singleobs` |
logical variable indicating whether a single observation should be retained for each subject |

`id` |
character string indicating which column of the data identifies subjects |

`time` |
character string indicating which column of the data contains the time at which the visit occurred |

`keep.first` |
logical variable indicating whether the first observation should be retained with probability 1. This is useful if the data consists of an observation at baseline followed by follow-up at stochastic time points. |

the outputted dataset.

Hoffman E, Sen P, Weinberg C. Within-cluster resampling. Biometrika 2001; 88:1121-1134

Follmann D, Proschan M, Leifer E. Multiple outputation: inference for complex clustered data by averaging analyses from independent data. Biometrics 2003; 59:420-429

Pullenayegum EM. Multiple outputation for the analysis of longitudinal data subject to irregular observation. Statistics in Medicine (in press).

Other mo: `mo`

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 | ```
library(nlme)
data(Phenobarb)
library(survival)
library(geepack)
Phenobarb$id <- as.numeric(Phenobarb$Subject)
Phenobarb$event <- as.numeric(is.finite(Phenobarb$conc))
Phenobarb.conc <- Phenobarb[is.finite(Phenobarb$conc),]
i <- iiw.weights(Surv(time.lag,time,event)~I(conc.lag>0) + conc.lag + cluster(Subject),
id="Subject",time="time",event="event",data=Phenobarb.conc,invariant="Subject",
lagvars=c("time","conc"),maxfu=NULL,lagfirst=0,first=TRUE)
Phenobarb.conc$weight <- i$iiw.weight
head(Phenobarb.conc)
data.output1 <- outputation(Phenobarb.conc,Phenobarb.conc$weight,singleobs=FALSE,
id="id",time="time",keep.first=FALSE)
head(data.output1)
data.output2 <- outputation(Phenobarb.conc,Phenobarb.conc$weight,singleobs=FALSE,
id="id",time="time",keep.first=FALSE)
head(data.output2)
data.output3 <- outputation(Phenobarb.conc,Phenobarb.conc$weight,singleobs=FALSE,
id="id",time="time",keep.first=FALSE)
head(data.output3)
# Note that the outputted dataset varies with each command run; outputation is done at random
``` |

IrregLong documentation built on May 2, 2019, 8:30 a.m.

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.