ci = bootci(nboot, bootfun, d)
ci = bootci(nboot, bootfun, d1, ..., dN)
ci = bootci(nboot, {bootfun, d1, ..., dN}, 'Type', type)
[ci, bootstat] = bootci(...)
| Parameter | Description |
|---|---|
| nboot | positive integer number of bootstrap samples. |
| bootfun | function handle applied to each bootstrap sample. |
| d | numeric or logical vector or matrix. Rows are sampled with replacement. |
| type | confidence interval type: 'bca', 'normal', 'percentile', or 'cper'. Studentized intervals are reserved for a future implementation. |
| Parameter | Description |
|---|---|
| ci | two-row array containing lower and upper confidence interval bounds. |
| bootstat | bootstrap statistics, one row per bootstrap sample. |
bootci draws bootstrap samples using Nelson's random generator and computes confidence intervals for statistics returned by a function handle.
rng('default');
ci = bootci(100, @mean, (1:10)', 'Type', 'percentile')