Description Usage Arguments Value Examples
View source: R/survBootOutliers.R
Extract the most outlying observations following a criteria based on the bootstrapped concordance with parallel processing
1 2 | survBootOutliers(surv.object, covariate.data, sod.method, B, B.N = NULL,
max.outliers, parallel.param = NULL)
|
surv.object |
An obect of type survival::Surv containing lifetimes and right-censoring status |
covariate.data |
A data frame containing the data with covariate values for each individual |
sod.method |
One of c("osd","bht","dbht","ld","martingale","deviance") |
B |
The number of bootstrap samples generated only applicable for "bht" and "dbht" methods. Typically at least 10x the size of the dataset, ideally should be increased until convergence. |
B.N |
the number of observations in each bootstrap sample |
max.outliers |
This parameter is only used for the "osd" method |
parallel.param |
(Optional) A BiocParallel object, examples: SerialParam(), MulticoreParam() |
For all methods except for "bht" and "dbht" the value returned is a data.frame containing the most outlying observations sorted by outlying score. For the "bht" method the value returned is a list of two members: "outlier_set": the most outlygin observations sorted by p-values; "histograms": histogram of concordance variation for each observation. For the "dbht" method the value returned is a list of two members: "outlier_set": the most outlygin observations sorted by p-values; "histograms": histogrms of concordance for each observations for the two types of bootstap: "poison" and "antidote".
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 | ## One Step Deletion "osd" method
## Not run:
whas <- get.whas100.dataset()
print( getwd() )
outliers_osd <- survBootOutliers(
surv.object=Surv(time = whas$times,event = whas$status ),
covariate.data = whas[,2:5],
sod.method = "osd",
max.outliers = 5
)
## End(Not run)
## Bootstrap Hypothesis Test "bht" with 1000 bootstrap samples,
## each with 100 individuals, running without parallelism.
## Not run:
whas <- get.whas100.dataset()
outliers_bht <- survBootOutliers(
surv.object=Surv(time = whas$times,event = whas$status ),
covariate.data = whas[,2:5],
sod.method = "bht",
B = 1000,
B.N = 100,
parallel.param = BiocParallel::MulticoreParam()
)
## End(Not run)
## Dual Bootstrap Hypothesis Test "dbht" with 1000 bootstrap samples,
## each with 50 individuals and running on all available cores.
## Not run: whas <- get.whas100.dataset()
outliers_dbht <- survBootOutliers(
surv.object=Surv(time = whas$times,event = whas$status ),
covariate.data = whas[,2:5],
sod.method = "dbht",
B = 1000,
B.N = 50,
parallel.param = BiocParallel::MulticoreParam()
)
## End(Not run)
## One Step Deletion "osd" with an amount of 10 for maximum outlier count
whas <- get.whas100.dataset()
outliers_osd <- survBootOutliers(
surv.object=Surv(time = whas$times,event = whas$status ),
covariate.data = whas[,2:5],
sod.method = "osd",
max.outliers = 10
)
## Likelihood displacement criterion for outlier ranking
whas <- get.whas100.dataset()
outliers_ld <- survBootOutliers(
surv.object=Surv(time = whas$times,event = whas$status ),
covariate.data = whas[,2:5],
sod.method = "ld"
)
## Cox regression deviance residuals criterion for outlier ranking
whas <- get.whas100.dataset()
outliers_deviance <- survBootOutliers(
surv.object=Surv(time = whas$times,event = whas$status ),
covariate.data = whas[,2:5],
sod.method = "deviance"
)
## Cox regression Martingale residuals criterion for outlier ranking
whas <- get.whas100.dataset()
outliers_martingale <- survBootOutliers(
surv.object=Surv(time = whas$times,event = whas$status ),
covariate.data = whas[,2:5],
sod.method = "martingale"
)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.