Skip to contents

Sample Nonparametric Curve Data

Source: R/nonparametric-curve-plot.R
sample_nonparametric_curve_data.Rd

Simulates pre-computed curve output matching what SAS produces after fitting a two-phase nonparametric temporal trend model and averaging patient-specific profiles with PROC SUMMARY. The output is suitable for direct use with hv_nonparametric().

Usage

sample_nonparametric_curve_data(
  n = 500,
  time_max = 12,
  n_points = 500,
  groups = NULL,
  outcome_type = c("probability", "continuous"),
  ci_level = 0.68,
  n_bins = 10,
  seed = 42L
)

Arguments

n

Number of simulated patients (used for binned data points and CI width). Default 500.

time_max

Upper end of the time axis (same units as the SAS iv_echo / iv_wristm variable). Default 12.

n_points

Number of time points on the fine prediction grid. Default 500 (matches the SAS inc=(max-min)/499.9 loop).

groups

NULL for a single average curve, or a named numeric vector of group-specific hazard multipliers, e.g. c("Ozaki" = 0.8, "CE-Pericardial" = 1.2).

outcome_type

"probability" (binary outcome, 0-1 scale) or "continuous". Default "probability".

ci_level

Confidence level for bootstrap-style CI bands. Default 0.68.

n_bins

Number of equal-sized data-summary bins. Default 10.

seed

Random seed. Default 42.

Value

A data frame with columns time, estimate, lower, upper, and (if groups is not NULL) group.

Details

SAS context: In the SAS templates this dataset corresponds to mean_curv (estimate column) plus boots_ci (lower/upper columns). Export those datasets to CSV and read them with read.csv() to use your own model output instead of this sample function.

See also

hv_nonparametric(), sample_nonparametric_curve_points()

Examples

# Single average curve
dat <- sample_nonparametric_curve_data(n = 500, time_max = 12)
head(dat)
#>         time  estimate     lower     upper
#> 1 0.05000000 0.2395042 0.2148392 0.2660414
#> 2 0.05055219 0.2395916 0.2149201 0.2661351
#> 3 0.05111048 0.2396798 0.2150019 0.2662297
#> 4 0.05167493 0.2397690 0.2150845 0.2663253
#> 5 0.05224562 0.2398592 0.2151680 0.2664220
#> 6 0.05282261 0.2399503 0.2152524 0.2665196

# Two-group comparison
dat2 <- sample_nonparametric_curve_data(
  n = 400, time_max = 7,
  groups = c("Ozaki" = 0.7, "CE-Pericardial" = 1.3),
  outcome_type = "continuous"
)
head(dat2)
#>         time estimate    lower    upper group
#> 1 0.05000000 40.42056 39.08636 41.75477 Ozaki
#> 2 0.05049761 40.42456 39.09036 41.75877 Ozaki
#> 3 0.05100018 40.42860 39.09439 41.76280 Ozaki
#> 4 0.05150775 40.43267 39.09846 41.76687 Ozaki
#> 5 0.05202037 40.43677 39.10257 41.77098 Ozaki
#> 6 0.05253809 40.44092 39.10671 41.77512 Ozaki