Example internal/external matched case-control individual-level data
ExampleData_cc_indi.RdA simulated 1:4 matched case-control dataset for illustrating
ncc_indi and cv.ncc_indi. The object contains
both an internal cohort and an external cohort in case-control form,
so that the external cohort can be used as individual-level external
information when fitting the conditional logistic regression model with
external integration.
This is the individual-level analog of ExampleData_cc_lowdim:
rather than supplying beta_external only, the external cohort is
provided in full (subject-level matched-set data).
Usage
data(ExampleData_cc_indi)Format
A list containing two cohorts:
- internal
A list with components for the internal matched case-control study:
- z
Numeric matrix of covariates of dimension \(1000\times 6\) with columns named
Z1–Z6.- y
Integer vector of binary outcomes (
1=case,0=control).- stratum
Integer vector identifying the matched set for each observation (200 unique strata, exactly one case and four controls per set).
- external
A list with the same component names as
internal, containing 500 matched sets (\(n = 2500\)) with the same 1:4 matching design, intended to be passed as the external cohort toncc_indi.
Details
Both the internal and external cohorts are 1:4 matched case-control studies
on six covariates Z1–Z6. The matching design (one case and
four controls per stratum) is preserved in both cohorts, which is required
by ncc_indi and the stratum-level fold construction in
cv.ncc_indi.
Examples
data(ExampleData_cc_indi)
str(ExampleData_cc_indi, max.level = 2)
#> List of 2
#> $ internal:List of 3
#> ..$ z : num [1:1000, 1:6] 1.4209 1.8271 1.2201 -0.0173 1.0239 ...
#> .. ..- attr(*, "dimnames")=List of 2
#> ..$ y : int [1:1000] 0 1 0 0 0 0 0 1 0 0 ...
#> ..$ stratum: int [1:1000] 1 1 1 1 1 2 2 2 2 2 ...
#> $ external:List of 3
#> ..$ z : num [1:2500, 1:6] -1.523 0.711 -0.674 -0.956 -0.805 ...
#> .. ..- attr(*, "dimnames")=List of 2
#> ..$ y : int [1:2500] 0 0 0 1 0 0 0 0 0 1 ...
#> ..$ stratum: int [1:2500] 1 1 1 1 1 2 2 2 2 2 ...