KMSubtractionError: Conduct Monte Carlo simulations to evaluate the limits of...

In josephjzhao/KMSubtraction: Reconstruction of unreported subgroup survival data utilizing published Kaplan-Meier survival curves

Description

This function conducts Monte Carlo simulations to evaluate the limits of error of KMSubtractionMatch() given parameters surrounding the reconstruction task required. Follow-up time was modeled by a random weibull distribution of common shape parameter of 1.000 and scale parameter of 5.000.

Usage

KMSubtractionError(
  n,
  censor_overall.p,
  censor_subgroup.p,
  subgroup.p,
  interval = 8,
  missing.p = 0,
  mc = 1000,
  ncores = 1,
  fig.label = ""
)

Arguments

n	Size of overall cohort.
censor_overall.p	Proportion of patients with censorship status in the overall cohort, which may be found from the object of KMSubtractionMatch().
censor_subgroup.p	Proportion of patients with censorship status in the subgroup cohort, which may be found from the object of KMSubtractionMatch().
subgroup.p	Proportion of reported subgroup.
interval	Number of number-at-risk table intervals.
missing.p	Proportion of missing data from the opposing subgroup. Default is set at 0.00.
mc	Number of Monte Carlo iterations. Default is set at 1000.
ncores	Number of cores to be utilized for parallel processing. The number of cores in your device may be found using the function 'parallel::detectCores()' from the parallel package.

Value

Reconstructed and original survival data were compared by means of marginal Cox-proportional hazard models and restricted mean survival time difference (RMSTD). This function returns density plots per matching algorithm for both ln(HR) and RMST-Difference; as well as summary statistics in table format.

Examples

Size of dataset, censorship proportion and subgroup proportion may be retrieved from the KMSubtractionMatch object, under Parameters.

data(cancer)
df_overall=colon
df_subgroup=colon[1:200,]
match=KMSubtractionMatch(df_overall, df_subgroup, matching="bipartite")

match$Parameters

n=match$Parameters[1,1]
subgroup.p=match$Parameters[3,1]
censor_overall.p=match$Parameters[4,1]
censor_subgroup.p=match$Parameters[5,1]

KMSubtractionError(n=n,
mc=1000,
censor_overall.p=censor_overall.p,
censor_subgroup.p=censor_subgroup.p,
subgroup.p=subgroup.p,
interval=8,
missing.p=0.02,
ncores=5)

josephjzhao/KMSubtraction documentation built on Jan. 25, 2022, 6:35 p.m.