diff --git a/README.md b/README.md index 2b710fae..d4e736f0 100755 --- a/README.md +++ b/README.md @@ -59,7 +59,7 @@ In Octave only, you can get a basic overview of the package and it's functions b ### Quick start -Below are some links to demonstrations of how the bootstrap or randomization functions from this package can be used to perform robust versions of commonly used statistical tests: +Below are some links to demonstrations of how the bootstrap or randomization functions from this package can be used to perform variants of the commonly used statistical tests below, but without the Normality assumption: * [t-test to compare two independent samples (assuming equal variances)](https://gnu-octave.github.io/statistics-resampling/function/boot1way.html#1) @@ -67,15 +67,15 @@ Below are some links to demonstrations of how the bootstrap or randomization fun * [t-test to compare two independent samples (not assuming equal variances)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#1) - * [t-test to compare two paired or matching samples (robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#2) + * [t-test to compare two paired or matching samples (also robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#2) - * [One-way ANOVA to compare more than two independent samples (robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#3) + * [One-way ANOVA to compare more than two independent samples (also robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#3) - * [Nested one-way ANOVA to compare more than two independent samples (robust to heteroscedasticity and grouping of observations)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#11) + * [Nested one-way ANOVA to compare more than two independent samples (also robust to heteroscedasticity and grouping of observations)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#11) - * [Two-way factorial ANOVA (robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#6) + * [Two-way factorial ANOVA (also robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#6) - * [One-way ANCOVA (robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#7) + * [One-way ANCOVA (also robust to heteroscedasticity)](https://gnu-octave.github.io/statistics-resampling/function/bootlm.html#7) ### Issues diff --git a/docs/function/boot1way.html b/docs/function/boot1way.html index f7879a89..76e5ed39 100644 --- a/docs/function/boot1way.html +++ b/docs/function/boot1way.html @@ -528,7 +528,7 @@

Demonstration 7

----------------------------------------------------------------------------- | Comparison | Test # | Ref # | Difference | t | p | |------------|------------|------------|------------|------------|----------| -| 1 | 2 | 1 | -0.9299 | -2.19 | .083 | +| 1 | 2 | 1 | +0.5927 | +1.34 | .192 | ----------------------------------------------------------------------------- | GROUP # | GROUP label | N | @@ -573,7 +573,7 @@

Demonstration 8

----------------------------------------------------------------------------- | Comparison | Test # | Ref # | Difference | t | p | |------------|------------|------------|------------|------------|----------| -| 1 | 2 | 1 | -0.6290 | -0.95 | .322 | +| 1 | 2 | 1 | +0.06620 | +0.14 | .820 | ----------------------------------------------------------------------------- | GROUP # | GROUP label | N | diff --git a/docs/function/bootclust.html b/docs/function/bootclust.html index f4bbcc2a..c1166a08 100644 --- a/docs/function/bootclust.html +++ b/docs/function/bootclust.html @@ -170,11 +170,11 @@

Demonstration 1

Resampling method: Balanced, bootstrap cluster resampling Number of resamples: 1999 Confidence interval (CI) type: Expanded bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 95% (1.0%, 97.2%) + Nominal coverage (and the percentiles used): 95% (1.1%, 97.2%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +29.65 -7.105e-15 +2.571 +23.47 +34.37 + +29.65 -3.197e-14 +2.564 +23.43 +34.48

Demonstration 2

@@ -200,11 +200,11 @@

Demonstration 2

Resampling method: Balanced, bootstrap cluster resampling Number of resamples: 1999 Confidence interval (CI) type: Expanded bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 95% (1.0%, 98.6%) + Nominal coverage (and the percentiles used): 95% (1.1%, 98.7%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +29.65 -0.02817 +2.924 +22.64 +35.92 + +29.65 -0.02999 +2.920 +22.64 +35.86

Demonstration 3

@@ -232,7 +232,7 @@

Demonstration 3

Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -6.414 +41.56 +97.48 +235.6 + +171.5 -6.379 +42.98 +96.61 +239.1

Demonstration 4

@@ -262,7 +262,7 @@

Demonstration 4

Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -9.188 +33.94 +104.4 +216.2 + +171.5 -9.524 +33.03 +106.0 +215.4

Demonstration 5

@@ -285,11 +285,11 @@

Demonstration 5

Resampling method: Balanced, bootstrap cluster resampling Number of resamples: 1999 Confidence interval (CI) type: Bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 90% (10.7%, 98.4%) + Nominal coverage (and the percentiles used): 90% (12.6%, 98.8%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -6.206 +41.36 +113.7 +256.0 + +171.5 -6.608 +42.24 +117.3 +267.8

Demonstration 6

@@ -314,11 +314,11 @@

Demonstration 6

Resampling method: Balanced, bootstrap cluster resampling Number of resamples: 1999 Confidence interval (CI) type: Bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 90% (12.7%, 98.6%) + Nominal coverage (and the percentiles used): 90% (13.7%, 98.8%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -9.686 +34.23 +121.6 +233.6 + +171.5 -9.512 +33.09 +125.0 +231.3

Demonstration 7

@@ -344,8 +344,8 @@

Demonstration 7

Bootstrap Statistics: original bias std_error CI_lower CI_upper - -0.07256 +0.01379 +0.2152 -0.4556 +0.2506 - -0.02682 +0.01024 +0.2075 -0.3029 +0.4062 + -0.09767 -0.02694 +0.2484 -0.4705 +0.3481 + -0.1495 -0.02999 +0.2569 -0.5599 +0.2306

Demonstration 8

@@ -372,8 +372,8 @@

Demonstration 8

Bootstrap Statistics: original bias std_error CI_lower CI_upper - +0.1431 +0.01975 +0.2204 -0.2553 +0.4740 - -0.4819 -0.003192 +0.1504 -0.6087 +0.1700 + -0.07584 -0.03771 +0.2730 -0.5805 +0.2926 + -0.05457 -0.05267 +0.3563 -0.6070 +0.5813

Demonstration 9

@@ -400,11 +400,11 @@

Demonstration 9

Resampling method: Balanced, bootstrap cluster resampling Number of resamples: 1999 Confidence interval (CI) type: Bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 95% (1.9%, 96.9%) + Nominal coverage (and the percentiles used): 95% (1.8%, 96.8%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +0.7764 -0.02490 +0.1436 +0.4011 +0.9949 + +0.7764 -0.02391 +0.1409 +0.4128 +0.9919

Package: statistics-resampling

diff --git a/docs/function/bootknife.html b/docs/function/bootknife.html index 32b96193..bf4a8a73 100644 --- a/docs/function/bootknife.html +++ b/docs/function/bootknife.html @@ -223,11 +223,11 @@

Demonstration 1

Number of resamples (outer): 1999 Number of resamples (inner): 0 Confidence interval (CI) type: Expanded bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 95% (1.4%, 97.4%) + Nominal coverage (and the percentiles used): 95% (1.3%, 97.2%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +29.65 -4.619e-14 +2.674 +23.66 +34.69 + +29.65 -2.842e-14 +2.695 +23.35 +34.56

Demonstration 2

@@ -253,11 +253,11 @@

Demonstration 2

Number of resamples (outer): 1999 Number of resamples (inner): 199 Confidence interval (CI) type: Calibrated percentile - Nominal coverage (and the percentiles used): 95% (1.1%, 97.2%) + Nominal coverage (and the percentiles used): 95% (0.9%, 98.1%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +29.65 +1.776e-14 +2.699 +23.53 +34.56 + +29.65 -1.208e-13 +2.797 +22.95 +35.05

Demonstration 3

@@ -284,11 +284,11 @@

Demonstration 3

Number of resamples (outer): 1999 Number of resamples (inner): 199 Confidence interval (CI) type: Calibrated percentile - Nominal coverage (and the percentiles used): 95% (2.4%, 98.2%) + Nominal coverage (and the percentiles used): 95% (3.0%, 97.7%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +30.86 -0.03345 +2.850 +24.65 +36.91 + +30.86 -0.006010 +2.849 +25.06 +36.65

Demonstration 4

@@ -317,7 +317,7 @@

Demonstration 4

Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -6.938 +42.95 +96.71 +237.8 + +171.5 -6.829 +43.16 +95.97 +237.3

Demonstration 5

@@ -341,11 +341,11 @@

Demonstration 5

Number of resamples (outer): 1999 Number of resamples (inner): 0 Confidence interval (CI) type: Bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 90% (13.3%, 98.9%) + Nominal coverage (and the percentiles used): 90% (13.1%, 98.9%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -6.545 +42.93 +117.7 +271.8 + +171.5 -7.292 +43.98 +114.3 +272.9

Demonstration 6

@@ -372,11 +372,11 @@

Demonstration 6

Number of resamples (outer): 1999 Number of resamples (inner): 199 Confidence interval (CI) type: Calibrated percentile (equal-tailed) - Nominal coverage (and the percentiles used): 90% (2.6%, 97.4%) + Nominal coverage (and the percentiles used): 90% (2.3%, 97.7%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -6.888 +45.39 +83.66 +250.1 + +171.5 -6.601 +45.18 +87.67 +255.0

Demonstration 7

@@ -402,11 +402,11 @@

Demonstration 7

Number of resamples (outer): 1999 Number of resamples (inner): 199 Confidence interval (CI) type: Calibrated percentile - Nominal coverage (and the percentiles used): 90% (12.1%, 99.5%) + Nominal coverage (and the percentiles used): 90% (12.9%, 99.5%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +171.5 -7.678 +44.20 +114.0 +277.0 + +171.5 -7.575 +44.09 +115.1 +283.2

Demonstration 8

@@ -433,8 +433,8 @@

Demonstration 8

Bootstrap Statistics: original bias std_error CI_lower CI_upper - +0.1674 -0.008330 +0.2082 -0.1522 +0.5412 - -0.06273 +0.03191 +0.3033 -0.5864 +0.3861 + +0.4165 -0.01155 +0.2633 +0.007875 +0.8868 + -0.3700 +0.01114 +0.2749 -0.8682 +0.04576

Demonstration 9

@@ -461,11 +461,11 @@

Demonstration 9

Number of resamples (outer): 1999 Number of resamples (inner): 0 Confidence interval (CI) type: Bias-corrected and accelerated (BCa) - Nominal coverage (and the percentiles used): 95% (0.5%, 93.5%) + Nominal coverage (and the percentiles used): 95% (0.6%, 93.8%) Bootstrap Statistics: original bias std_error CI_lower CI_upper - +0.7764 -0.007956 +0.1380 +0.3146 +0.9468 + +0.7764 -0.005985 +0.1371 +0.3509 +0.9502

Demonstration 10

@@ -577,13 +577,13 @@

Demonstration 10

Bootstrap Statistics: original bias std_error CI_lower CI_upper - -16.69 -0.5227 +2.108 -20.27 -12.48 - -11.72 -0.3645 +1.850 -14.98 -8.003 - -8.061 -0.2629 +1.756 -11.25 -4.436 - +0.1048 +0.005499 +0.09200 -0.08304 +0.2818 - +0.01034 +0.0007184 +0.006770 -0.001595 +0.02395 - +0.06452 +0.002318 +0.01672 +0.03116 +0.09602 - +0.001664 +6.690e-06 +0.0007557 +0.0001953 +0.003135 + -16.69 -0.5168 +2.137 -20.46 -12.24 + -11.72 -0.3592 +1.862 -15.08 -7.883 + -8.061 -0.2570 +1.768 -11.39 -4.467 + +0.1048 +0.005554 +0.09117 -0.07452 +0.2804 + +0.01034 +0.0007293 +0.006650 -0.002291 +0.02328 + +0.06452 +0.002379 +0.01689 +0.03089 +0.09543 + +0.001664 +1.194e-06 +0.0007440 +0.0002257 +0.003140

Demonstration 11

diff --git a/docs/function/bootmode.html b/docs/function/bootmode.html index 7ebf7ecc..c4aa9ded 100644 --- a/docs/function/bootmode.html +++ b/docs/function/bootmode.html @@ -151,7 +151,7 @@

Demonstration 1

ans = H1 is 1 with p = 0.001 so reject the null hypothesisthat there is 1 mode -ans = H2 is 0 with p = 0.332 so accept the null hypothesis that there are 2 modes +ans = H2 is 0 with p = 0.327 so accept the null hypothesis that there are 2 modes

Package: statistics-resampling

diff --git a/docs/function/images/boot1way_701.png b/docs/function/images/boot1way_701.png index 5ed7f90e..9a5a8f91 100644 Binary files a/docs/function/images/boot1way_701.png and b/docs/function/images/boot1way_701.png differ diff --git a/docs/function/images/boot1way_801.png b/docs/function/images/boot1way_801.png index a1903e53..b2094976 100644 Binary files a/docs/function/images/boot1way_801.png and b/docs/function/images/boot1way_801.png differ diff --git a/docs/readme.html b/docs/readme.html index 386b21f5..f1d59ecf 100644 --- a/docs/readme.html +++ b/docs/readme.html @@ -123,8 +123,9 @@

Usage

href="https://octave.discourse.group/c/help/6">discourse group.

Quick start

Below are some links to demonstrations of how the bootstrap or -randomization functions from this package can be used to perform robust -versions of commonly used statistical tests:

+randomization functions from this package can be used to perform +variants of the commonly used statistical tests below, but without the +Normality assumption:

Issues

If you find bugs or have any suggestions, please raise an issue on