BootstrapResult

Result of a bootstrap confidence interval computation.

Usage

BootstrapResult(
    estimate,
    se,
    conf_low,
    conf_high,
    conf_level,
    ci_type,
    n_boot,
    distribution,
)

Holds the point estimate, bootstrap standard error, and confidence interval for a Kaplan-Meier summary statistic. The full bootstrap distribution is retained for diagnostics or custom interval construction.

Attributes

estimate: float: Point estimate from the original (non-resampled) data.
se: float: Bootstrap standard error (standard deviation of the bootstrap distribution).
conf_low: float: Lower confidence limit.
conf_high: float: Upper confidence limit.
conf_level: float: Confidence level used for the interval.
ci_type: str: Type of confidence interval ("percentile", "normal", or "bca").
n_boot: int: Number of bootstrap replicates.
distribution: Array: The full bootstrap distribution (array of length n_boot=).

Examples

Bootstrap a 95% confidence interval for the median survival time on the lung dataset:

import greenwood as gw

lung = gw.load_dataset("lung", backend="polars")
death = gw.Outcome.surv(time="time", event="status")

result = gw.bootstrap(death, statistic="median", n_boot=500, seed=0, data=lung)
result

BootstrapResult (n_boot=500, ci_type='percentile')

  estimate      se  0.95LCL  0.95UCL
   310.000  23.330  283.475  363.000

Export as a single-row DataFrame:

result.to_frame(format="polars")

shape: (1, 4)

estimate	se	conf_low	conf_high
f64	f64	f64	f64
310.0	23.330162	283.475	363.0

Methods

Name	Description
to_frame()	Export the result as a single-row DataFrame.

to_frame()

Export the result as a single-row DataFrame.

Usage

Source

to_frame(
    *,
    format=None,
)

Parameters

format: str | None = None: Backend format: "pandas", "polars", or "pyarrow". Default (None) uses the first available backend.

Examples

import greenwood as gw

lung = gw.load_dataset("lung", backend="polars")
death = gw.Outcome.surv(time="time", event="status")

result = gw.bootstrap(death, statistic="median", n_boot=500, seed=0, data=lung)
result.to_frame(format="polars")

shape: (1, 4)

estimate	se	conf_low	conf_high
f64	f64	f64	f64
310.0	23.330162	283.475	363.0