Description Usage Arguments Details Value Author(s) See Also Examples
View source: R/interim_and_survival_package_functions.R
interimTrial comes in two variants: When
standalone
is 'TRUE' the result is nicely
formatted and can be used with the plotting functions
plotinterimFDR
, plotinterimAPR
, and
plotinterimStops
. When standalone
is
'FALSE' the result is presented as one single vector,
which makes it usable within
interimTrialSimulation
.
1 2 | interimTrial(patdat, gendat, l.2, M.1, M.2, alpha,
powerThreshold, adjustment, standalone = TRUE)
|
patdat |
the patient data (arrival times and
survival times) as generated by
|
gendat |
the gene expression data as generated by
|
l.2 |
the length of the follow-up period to be simulated |
M.1 |
the number of analyses during the recruitment period to be simulated |
M.2 |
the number of analyses during the follow-up period to be simulated |
alpha |
the error level at which the FDR is to be controlled |
powerThreshold |
the study is stopped as soon as the estimated power rate exceeds this threshold |
adjustment |
the method to use for the p-value adjustment to account for multiple testing. In c("holm", "hochberg", "hommel", "bonferroni", "BH", "BY", "fdr", "none") |
standalone |
boolean parameter that controls whether it is used within a bigger simulation. Defaults to 'TRUE' |
The adjustment
parameter is used as parameter to
p.adjust
.
The parameters patdat
and gendat
are of
type list
and do not necessarily have to be
obtained from the functions generatePatientData
and generateExpressionData
. Especially the
entries N.1
and N.2
in patdat
can be
changed. But in this case make sure to generate
gendat
accordingly.
Internally interimTrial performs a gene-wise Cox-regression to determin the survival correlated genes.
In case the function is used in standalone mode:
resultTable |
the only real result. This table holds the error and power rates that came from the simulation. The fist row additionally shows at which analysis the simulated survival study was stopped. |
The other values are copied versions of the corresponding input parameters. They are included into the result for convenience only.
In case the function is not used in standalone mode: The
return value is one vector containing the elements of
resultTable
.
Andreas Leha andreas.leha@med.uni-goettingen.de
generatePatientData
and
generateExpressionData
to generate the
input data
plotinterimFDR
,
plotinterimAPR
, and
plotinterimStops
to visualize the results
interimTrialSimulation
for a wrapper that
simulates not one but many survival studies (and calls
this function)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 | N <- 50
l.1.tick <- 60
lambda <- 60
patdat <- generatePatientData(N, l.1.tick, lambda)
fc <- c(rep(0, 500), ceiling(rnorm(500, 0, 1)-0.5))
Sigma.1 <- diag(1000)
Sigma.2 <- diag(1000)
N.1 <- patdat$N.1
N.2 <- patdat$N.2
gendat <- generateExpressionData(fc, Sigma.1, Sigma.2, N.1, N.2)
l.2 <- 60
M.1 <- 2
M.2 <- 2
alpha <- 0.05
powerThreshold <- 0.8
adjustment <- "BH"
## Not run: T <- interimTrial(patdat, gendat, l.2, M.1, M.2, alpha, powerThreshold, adjustment)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.