2020< div id ="doccontent ">
2121
2222< h2 > bootint</ h2 >
23- < pre > Computes simple percentile confidence interval(s) directly from a vector (or
24- row- major matrix) of bootstrap statistics.
23+ < pre > Computes percentile confidence interval(s) directly from a vector (or row-
24+ major matrix) of bootstrap statistics.
2525
2626 -- Function File: CI = bootint (BOOTSTAT)
2727 -- Function File: CI = bootint (BOOTSTAT, PROB)
28+ -- Function File: CI = bootint (BOOTSTAT, PROB, ORIGINAL)
2829
29- 'CI = bootint (BOOTSTAT)' computes 95% percentile confidence intervals
30- directly from the vector, or rows* of the matrix in BOOTSTAT, where
31- BOOTSTAT contains bootstrap statistics such as those generated using
32- the `bootstrp` function. Depending on the application, bootstrap
33- confidence intervals with better coverage and accuracy can be computed
34- using the various dedicated bootstrap functions from the statistics-
35- resampling package.
30+ 'CI = bootint (BOOTSTAT)' computes simple 95% percentile confidence
31+ intervals [1,2] directly from the vector, or rows* of the matrix in
32+ BOOTSTAT, where BOOTSTAT contains bootstrap statistics such as those
33+ generated using the `bootstrp` function. Depending on the application,
34+ bootstrap confidence intervals with better coverage and accuracy can
35+ be computed using the various dedicated bootstrap confidence interval
36+ functions from the statistics- resampling package.
3637
3738 * The matrix should have dimensions P * NBOOT, where P corresponds to
3839 the number of parameter estimates and NBOOT corresponds to the number
@@ -50,7 +51,22 @@ <h2>bootint</h2>
5051 The default value of PROB is the vector: [0.025, 0.975], for an
5152 equal-tailed 95% percentile confidence interval.
5253
53- bootint (version 2024.04.24)
54+ 'CI = bootint (BOOTSTAT, PROB, ORIGINAL)' uses the ORIGINAL estimates
55+ associated with BOOTSTAT to correct PROB and the resulting confidence
56+ intervals (CI) for median bias. The confidence intervals returned in CI
57+ therefore become bias-corrected percentile intervals [3,4].
58+
59+ BIBLIOGRAPHY:
60+ [1] Efron (1979) Bootstrap Methods: Another look at the jackknife.
61+ Annals Stat. 7,1-26
62+ [2] Efron, and Tibshirani (1993) An Introduction to the Bootstrap.
63+ New York, NY: Chapman & Hall
64+ [3] Efron (1981) Nonparametric Standard Errors and Confidence Intervals.
65+ Can J Stat. 9(2):139-172
66+ [4] Efron (1982) The jackknife, the bootstrap, and other resampling plans.
67+ SIAM-NSF, CBMS #38
68+
69+ bootint (version 2024.05.19)
5470 Author: Andrew Charles Penn
5571 https://www.researchgate.net/profile/Andrew_Penn/
5672
@@ -69,6 +85,35 @@ <h2>bootint</h2>
6985 along with this program. If not, see http://www.gnu.org/licenses/
7086</ pre >
7187
88+ < h3 > < a name ="1 "> Demonstration 1</ a > </ h3 >
89+ < div class ="demo ">
90+ < p > The following code</ p >
91+ < pre class ="example ">
92+
93+ % Law school data
94+ data = [576, 3.39; 635, 3.30; 558, 2.81; 578, 3.03; 666, 3.44; ...
95+ 580, 3.07; 555, 3.00; 661, 3.43; 661, 3.36; 605, 3.13; ...
96+ 653, 3.12; 575, 2.74; 545, 2.76; 572, 2.88; 594, 2.96];
97+ x = data(:, 1);
98+ y = data(:, 2);
99+ r = cor (x, y);
100+
101+ % 95% confidence interval for the mean
102+ bootstat = bootstrp (4999, @cor, x, y);
103+ CI_per = bootint (bootstat,0.95) % 95% simple percentile interval
104+ CI_cper = bootint (bootstat,0.95,r) % 95% bias-corrected percentile interval
105+
106+ % Please be patient, the calculations will be completed soon...</ pre >
107+ < p > Produces the following output</ p >
108+ < pre class ="example "> CI_per =
109+
110+ 0.45985 0.96204
111+
112+ CI_cper =
113+
114+ 0.41869 0.95609</ pre >
115+ </ div >
116+
72117< p > Package: < a href ="../index.html "> statistics-resampling</ a > </ p >
73118</ div >
74119</ body >
0 commit comments