This page contains full statistical result of the paper “MOAD: Modeling Observation-based Approximate Depedency”, in the proceedings of the 19th IEEE international Working Conference on Source Code Analysis and Manipulation (SCAM 2019).
slice_size <- read.csv("data/slice_size.csv")
summary(slice_size)## X proj crit scheme
## Min. : 0.0 replace :1518 Min. : 1.00 2hot_0 :2700
## 1st Qu.: 76.0 totinfo :1260 1st Qu.: 26.00 onehot_0:2700
## Median :173.0 printtokens : 588 Median : 58.00
## Mean :233.0 schedule2 : 486 Mean : 78.32
## 3rd Qu.:356.2 printtokens2: 450 3rd Qu.:119.25
## Max. :758.0 schedule : 450 Max. :253.00
## (Other) : 648
## infer size
## logistic :1800 Min. :0.06202
## once_success:1800 1st Qu.:0.42210
## simple_bayes:1800 Median :0.52756
## Mean :0.52483
## 3rd Qu.:0.63327
## Max. :0.92580
## onehot_size_data <- slice_size[slice_size$scheme=='onehot_0',]
anova1 <- aov(size ~ infer, data = onehot_size_data)
summary(anova1)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 6.53 3.263 217.7 <2e-16 ***
## Residuals 2697 40.43 0.015
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova1, 'infer', group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 0.01498974 2697 0.5680238 21.55413 0.01353422
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer 3 3.316328 0.05
##
## $means
## size std r Min Max Q25
## logistic 0.6315512 0.09280558 900 0.3387097 0.9257951 0.5686901
## once_success 0.5117917 0.14759745 900 0.1136951 0.9257951 0.4002892
## simple_bayes 0.5607285 0.12071174 900 0.2459016 0.9257951 0.4664537
## Q50 Q75
## logistic 0.6299213 0.6823956
## once_success 0.4814815 0.6102236
## simple_bayes 0.5527157 0.6329775
##
## $comparison
## NULL
##
## $groups
## size groups
## logistic 0.6315512 a
## simple_bayes 0.5607285 b
## once_success 0.5117917 c
##
## attr(,"class")
## [1] "group"twohot_size_data <- slice_size[slice_size$scheme=='2hot_0',]
anova2 <- aov(size ~ infer, data = twohot_size_data)
summary(anova2)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 8.88 4.440 173.5 <2e-16 ***
## Residuals 2697 69.03 0.026
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova2, 'infer', group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 0.02559619 2697 0.481633 33.21784 0.01768576
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer 3 3.316328 0.05
##
## $means
## size std r Min Max Q25
## logistic 0.4644182 0.1771996 900 0.1487805 0.9257951 0.3481386
## once_success 0.4216035 0.1786931 900 0.0620155 0.9257951 0.2834646
## simple_bayes 0.5588774 0.1160072 900 0.2741935 0.9257951 0.4746377
## Q50 Q75
## logistic 0.4440895 0.5902439
## once_success 0.3850129 0.5826772
## simple_bayes 0.5467850 0.6134185
##
## $comparison
## NULL
##
## $groups
## size groups
## simple_bayes 0.5588774 a
## logistic 0.4644182 b
## once_success 0.4216035 c
##
## attr(,"class")
## [1] "group"a_both <- aov(size ~ infer * scheme, data = slice_size)
summary(a_both)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 9.25 4.625 227.9 <2e-16 ***
## scheme 1 10.08 10.076 496.5 <2e-16 ***
## infer:scheme 2 6.16 3.078 151.7 <2e-16 ***
## Residuals 5394 109.46 0.020
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(a_both, c('infer', 'scheme'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 0.02029296 5394 0.5248284 27.14285 0.01914357
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:scheme 6 4.031544 0.05
##
## $means
## size std r Min Max
## logistic:2hot_0 0.4644182 0.17719961 900 0.1487805 0.9257951
## logistic:onehot_0 0.6315512 0.09280558 900 0.3387097 0.9257951
## once_success:2hot_0 0.4216035 0.17869305 900 0.0620155 0.9257951
## once_success:onehot_0 0.5117917 0.14759745 900 0.1136951 0.9257951
## simple_bayes:2hot_0 0.5588774 0.11600715 900 0.2741935 0.9257951
## simple_bayes:onehot_0 0.5607285 0.12071174 900 0.2459016 0.9257951
## Q25 Q50 Q75
## logistic:2hot_0 0.3481386 0.4440895 0.5902439
## logistic:onehot_0 0.5686901 0.6299213 0.6823956
## once_success:2hot_0 0.2834646 0.3850129 0.5826772
## once_success:onehot_0 0.4002892 0.4814815 0.6102236
## simple_bayes:2hot_0 0.4746377 0.5467850 0.6134185
## simple_bayes:onehot_0 0.4664537 0.5527157 0.6329775
##
## $comparison
## NULL
##
## $groups
## size groups
## logistic:onehot_0 0.6315512 a
## simple_bayes:onehot_0 0.5607285 b
## simple_bayes:2hot_0 0.5588774 b
## once_success:onehot_0 0.5117917 c
## logistic:2hot_0 0.4644182 d
## once_success:2hot_0 0.4216035 e
##
## attr(,"class")
## [1] "group"miss_excess <- read.csv("data/miss_excess.csv")
summary(miss_excess)## X.1 X doe infer
## Min. : 12961 Min. : 10530 2hot_0 :8037 logistic :5358
## 1st Qu.:390281 1st Qu.: 61510 onehot_0:8037 once_success:5358
## Median :394300 Median :169568 simple_bayes:5358
## Mean :439261 Mean :127238
## 3rd Qu.:696362 3rd Qu.:205465
## Max. :754362 Max. :209483
##
## idx level common excess
## Min. : 1.00 tline:5358 Min. : 0.0 Min. : 0.00
## 1st Qu.: 26.00 tstmt:5358 1st Qu.: 54.0 1st Qu.: 3.00
## Median : 58.00 wline:5358 Median :109.0 Median : 12.00
## Mean : 78.65 Mean :120.6 Mean : 28.83
## 3rd Qu.:120.00 3rd Qu.:161.0 3rd Qu.: 42.00
## Max. :253.00 Max. :392.0 Max. :324.00
##
## miss sim succ proj
## Min. : 0.00 Min. :0.0000 False:10338 replace :4554
## 1st Qu.: 27.00 1st Qu.:0.3655 True : 5736 totinfo :3762
## Median : 53.00 Median :0.5637 printtokens :1764
## Mean : 63.98 Mean :0.5443 schedule2 :1386
## 3rd Qu.: 88.00 3rd Qu.:0.7446 printtokens2:1350
## Max. :333.00 Max. :0.9805 schedule :1314
## (Other) :1944anova_miss <- aov(miss ~ infer, data = miss_excess)
summary(anova_miss)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 3853187 1926594 899.8 <2e-16 ***
## Residuals 16071 34409091 2141
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova_miss, 'infer', group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 2141.067 16071 63.98252 72.31923 2.095421
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer 3 3.314801 0.05
##
## $means
## miss std r Min Max Q25 Q50 Q75
## logistic 75.52613 54.54067 5358 0 333 36 65 100
## once_success 42.09761 30.68412 5358 0 170 18 35 61
## simple_bayes 74.32381 50.06997 5358 0 244 36 67 103
##
## $comparison
## NULL
##
## $groups
## miss groups
## logistic 75.52613 a
## simple_bayes 74.32381 a
## once_success 42.09761 b
##
## attr(,"class")
## [1] "group"me_W_line <- miss_excess[miss_excess$level == 'wline',]
anova <- aov(miss ~ infer * doe * proj, data = me_W_line)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 1045585 522792 597.94 <2e-16 ***
## doe 1 802969 802969 918.38 <2e-16 ***
## proj 9 4358385 484265 553.87 <2e-16 ***
## infer:doe 2 449810 224905 257.23 <2e-16 ***
## infer:proj 18 591003 32833 37.55 <2e-16 ***
## doe:proj 9 159021 17669 20.21 <2e-16 ***
## infer:doe:proj 18 128887 7160 8.19 <2e-16 ***
## Residuals 5298 4632193 874
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, c('infer', 'doe'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 874.3286 5298 78.63792 37.60151 3.989201
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:doe 6 4.031571 0.05
##
## $means
## miss std r Min Max Q25 Q50 Q75
## logistic:2hot_0 66.90146 32.14934 893 6 172 42 62 91
## logistic:onehot_0 111.31019 62.89789 893 6 329 70 96 143
## once_success:2hot_0 44.45801 20.50478 893 2 105 30 41 60
## once_success:onehot_0 73.33035 33.07421 893 6 170 50 71 98
## simple_bayes:2hot_0 87.82867 47.58281 893 5 244 54 83 112
## simple_bayes:onehot_0 87.99888 47.91923 893 6 240 56 84 111
##
## $comparison
## NULL
##
## $groups
## miss groups
## logistic:onehot_0 111.31019 a
## simple_bayes:onehot_0 87.99888 b
## simple_bayes:2hot_0 87.82867 b
## once_success:onehot_0 73.33035 c
## logistic:2hot_0 66.90146 d
## once_success:2hot_0 44.45801 e
##
## attr(,"class")
## [1] "group"me_T_line <- miss_excess[miss_excess$level == 'tline',]
anova <- aov(miss ~ infer * doe * proj, data = me_T_line)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 1842399 921200 770.963 < 2e-16 ***
## doe 1 820571 820571 686.745 < 2e-16 ***
## proj 9 3695552 410617 343.650 < 2e-16 ***
## infer:doe 2 414518 207259 173.458 < 2e-16 ***
## infer:proj 18 726220 40346 33.766 < 2e-16 ***
## doe:proj 9 154432 17159 14.361 < 2e-16 ***
## infer:doe:proj 18 97191 5399 4.519 6.1e-10 ***
## Residuals 5298 6330415 1195
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, c('infer', 'doe'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 1194.869 5298 63.46491 54.46612 4.663464
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:doe 6 4.031571 0.05
##
## $means
## miss std r Min Max Q25 Q50 Q75
## logistic:2hot_0 57.44569 34.31192 893 2 170 31 48 86
## logistic:onehot_0 100.41433 66.74735 893 2 333 50 80 140
## once_success:2hot_0 22.23404 12.01009 893 1 72 14 21 27
## once_success:onehot_0 52.54759 29.09292 893 2 159 31 49 74
## simple_bayes:2hot_0 73.58903 53.61656 893 2 241 35 60 105
## simple_bayes:onehot_0 74.55879 53.27493 893 2 237 36 65 99
##
## $comparison
## NULL
##
## $groups
## miss groups
## logistic:onehot_0 100.41433 a
## simple_bayes:onehot_0 74.55879 b
## simple_bayes:2hot_0 73.58903 b
## logistic:2hot_0 57.44569 c
## once_success:onehot_0 52.54759 d
## once_success:2hot_0 22.23404 e
##
## attr(,"class")
## [1] "group"me_T_stmt <- miss_excess[miss_excess$level == 'tstmt',]
anova <- aov(miss ~ infer * doe * proj, data = me_T_stmt)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 1059513 529757 520.527 < 2e-16 ***
## doe 1 675840 675840 664.065 < 2e-16 ***
## proj 9 1804725 200525 197.031 < 2e-16 ***
## infer:doe 2 321280 160640 157.841 < 2e-16 ***
## infer:proj 18 352581 19588 19.247 < 2e-16 ***
## doe:proj 9 111476 12386 12.170 < 2e-16 ***
## infer:doe:proj 18 72592 4033 3.963 3.11e-08 ***
## Residuals 5298 5391944 1018
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, c('infer', 'doe'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 1017.732 5298 49.84472 64.00259 4.303931
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:doe 6 4.031571 0.05
##
## $means
## miss std r Min Max Q25 Q50 Q75
## logistic:2hot_0 40.19149 25.30022 893 0 136 22 33 60
## logistic:onehot_0 76.89362 55.19823 893 0 290 32 66 119
## once_success:2hot_0 15.13102 10.26945 893 0 51 7 13 21
## once_success:onehot_0 44.88466 28.66435 893 0 129 23 41 64
## simple_bayes:2hot_0 60.51848 45.02596 893 0 209 25 52 92
## simple_bayes:onehot_0 61.44905 45.03610 893 0 208 25 52 94
##
## $comparison
## NULL
##
## $groups
## miss groups
## logistic:onehot_0 76.89362 a
## simple_bayes:onehot_0 61.44905 b
## simple_bayes:2hot_0 60.51848 b
## once_success:onehot_0 44.88466 c
## logistic:2hot_0 40.19149 d
## once_success:2hot_0 15.13102 e
##
## attr(,"class")
## [1] "group"anova_excess <- aov(excess ~ infer, data = miss_excess)
summary(anova_excess)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 164377 82188 58.62 <2e-16 ***
## Residuals 16071 22530816 1402
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova_excess, 'infer', group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 1401.955 16071 28.83483 129.8523 1.695599
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer 3 3.314801 0.05
##
## $means
## excess std r Min Max Q25 Q50 Q75
## logistic 33.15398 39.67437 5358 0 324 4 16 52
## once_success 27.83632 38.31679 5358 0 257 3 11 38
## simple_bayes 25.51418 34.11206 5358 0 184 3 9 38
##
## $comparison
## NULL
##
## $groups
## excess groups
## logistic 33.15398 a
## once_success 27.83632 b
## simple_bayes 25.51418 c
##
## attr(,"class")
## [1] "group"me_W_line <- miss_excess[miss_excess$level == 'wline',]
anova <- aov(excess ~ infer * doe * proj, data = me_W_line)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 34152 17076 28.270 6.13e-13 ***
## doe 1 70018 70018 115.918 < 2e-16 ***
## proj 9 6400049 711117 1177.277 < 2e-16 ***
## infer:doe 2 70797 35399 58.604 < 2e-16 ***
## infer:proj 18 121587 6755 11.183 < 2e-16 ***
## doe:proj 9 40484 4498 7.447 6.89e-11 ***
## infer:doe:proj 18 97418 5412 8.960 < 2e-16 ***
## Residuals 5298 3200177 604
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, c('infer', 'doe'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 604.0349 5298 58.3156 42.14502 3.315733
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:doe 6 4.031571 0.05
##
## $means
## excess std r Min Max Q25 Q50 Q75
## logistic:2hot_0 68.33259 48.31075 893 0 324 28 64 94
## logistic:onehot_0 51.10078 39.09685 893 0 176 24 37 69
## once_success:2hot_0 62.60470 47.54987 893 0 228 30 47 76
## once_success:onehot_0 58.31355 41.81081 893 0 199 29 46 76
## simple_bayes:2hot_0 54.85442 39.97560 893 0 181 27 44 67
## simple_bayes:onehot_0 54.68757 39.79721 893 0 184 27 44 68
##
## $comparison
## NULL
##
## $groups
## excess groups
## logistic:2hot_0 68.33259 a
## once_success:2hot_0 62.60470 b
## once_success:onehot_0 58.31355 c
## simple_bayes:2hot_0 54.85442 d
## simple_bayes:onehot_0 54.68757 d
## logistic:onehot_0 51.10078 e
##
## attr(,"class")
## [1] "group"me_T_line <- miss_excess[miss_excess$level == 'tline',]
anova <- aov(excess ~ infer * doe * proj, data = me_T_line)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 113158 56579 158.98 <2e-16 ***
## doe 1 64944 64944 182.48 <2e-16 ***
## proj 9 838772 93197 261.87 <2e-16 ***
## infer:doe 2 94529 47264 132.80 <2e-16 ***
## infer:proj 18 151357 8409 23.63 <2e-16 ***
## doe:proj 9 41898 4655 13.08 <2e-16 ***
## infer:doe:proj 18 89002 4945 13.89 <2e-16 ***
## Residuals 5298 1885528 356
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, c('infer', 'doe'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 355.8943 5298 18.26017 103.3132 2.545125
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:doe 6 4.031571 0.05
##
## $means
## excess std r Min Max Q25 Q50 Q75
## logistic:2hot_0 33.99440 36.45833 893 0 240 7 17 63
## logistic:onehot_0 15.32251 16.63224 893 0 145 6 10 19
## once_success:2hot_0 15.49832 24.55052 893 0 257 4 8 18
## once_success:onehot_0 12.64838 18.95237 893 0 192 4 6 14
## simple_bayes:2hot_0 15.73236 19.82092 893 0 156 4 6 24
## simple_bayes:onehot_0 16.36506 20.24721 893 0 156 4 7 25
##
## $comparison
## NULL
##
## $groups
## excess groups
## logistic:2hot_0 33.99440 a
## simple_bayes:onehot_0 16.36506 b
## simple_bayes:2hot_0 15.73236 b
## once_success:2hot_0 15.49832 b
## logistic:onehot_0 15.32251 b
## once_success:onehot_0 12.64838 c
##
## attr(,"class")
## [1] "group"me_T_stmt <- miss_excess[miss_excess$level == 'tstmt',]
anova <- aov(excess ~ infer * doe * proj, data = me_T_stmt)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## infer 2 80730 40365 141.74 <2e-16 ***
## doe 1 79463 79463 279.02 <2e-16 ***
## proj 9 200322 22258 78.16 <2e-16 ***
## infer:doe 2 86894 43447 152.56 <2e-16 ***
## infer:proj 18 109362 6076 21.33 <2e-16 ***
## doe:proj 9 45025 5003 17.57 <2e-16 ***
## infer:doe:proj 18 99679 5538 19.45 <2e-16 ***
## Residuals 5298 1508802 285
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, c('infer', 'doe'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 284.7871 5298 9.928705 169.9682 2.276715
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:doe 6 4.031571 0.05
##
## $means
## excess std r Min Max Q25 Q50 Q75
## logistic:2hot_0 24.510638 33.49214 893 0 215 2 7 50
## logistic:onehot_0 5.662934 12.30077 893 0 116 1 2 5
## once_success:2hot_0 11.058231 20.39330 893 0 229 2 5 12
## once_success:onehot_0 6.894737 13.64032 893 0 154 2 3 7
## simple_bayes:2hot_0 5.770437 12.83755 893 0 143 1 3 5
## simple_bayes:onehot_0 5.675252 12.69272 893 0 143 1 2 5
##
## $comparison
## NULL
##
## $groups
## excess groups
## logistic:2hot_0 24.510638 a
## once_success:2hot_0 11.058231 b
## once_success:onehot_0 6.894737 c
## simple_bayes:2hot_0 5.770437 c
## simple_bayes:onehot_0 5.675252 c
## logistic:onehot_0 5.662934 c
##
## attr(,"class")
## [1] "group"sample_size <- read.csv("data/sample.csv")
summary(sample_size)## X proj crit scheme
## Min. : 0.0 replace :7590 Min. : 1.00 2hot_.1_0: 2700
## 1st Qu.: 76.0 totinfo :6300 1st Qu.: 26.00 2hot_.2_0: 2700
## Median :173.0 printtokens :2940 Median : 58.00 2hot_.3_0: 2700
## Mean :233.0 schedule2 :2430 Mean : 78.32 2hot_.4_0: 2700
## 3rd Qu.:356.2 printtokens2:2250 3rd Qu.:119.25 2hot_.5_0: 2700
## Max. :758.0 schedule :2250 Max. :253.00 2hot_.6_0: 2700
## (Other) :3240 (Other) :10800
## infer size
## logistic :9000 Min. :0.06202
## once_success:9000 1st Qu.:0.39617
## simple_bayes:9000 Median :0.50160
## Mean :0.50202
## 3rd Qu.:0.61342
## Max. :0.92580
## sample_size_O <- sample_size[sample_size$infer == 'once_success',]
anova <- aov(size ~ scheme, data = sample_size_O)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## scheme 9 3.61 0.4006 13.65 <2e-16 ***
## Residuals 8990 263.85 0.0293
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, 'scheme', group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 0.02934889 8990 0.4456162 38.44456 0.02555597
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey scheme 10 4.475256 0.05
##
## $means
## size std r Min Max Q25 Q50
## 2hot_.1_0 0.4880421 0.1587421 900 0.10335917 0.9257951 0.3731884 0.4520241
## 2hot_.2_0 0.4697844 0.1648237 900 0.10077519 0.9257951 0.3562992 0.4440895
## 2hot_.3_0 0.4581279 0.1691248 900 0.08536585 0.9257951 0.3322684 0.4237726
## 2hot_.4_0 0.4486385 0.1708800 900 0.08010336 0.9257951 0.3226837 0.4121406
## 2hot_.5_0 0.4435669 0.1712949 900 0.07751938 0.9257951 0.3180212 0.4089457
## 2hot_.6_0 0.4370940 0.1725848 900 0.07751938 0.9257951 0.3041868 0.4057508
## 2hot_.7_0 0.4340090 0.1740101 900 0.07235142 0.9257951 0.2966889 0.4057508
## 2hot_.8_0 0.4309716 0.1750216 900 0.06829268 0.9257951 0.2962963 0.3984610
## 2hot_.9_0 0.4243241 0.1770567 900 0.06201550 0.9257951 0.2907348 0.3851894
## 2hot_0 0.4216035 0.1786931 900 0.06201550 0.9257951 0.2834646 0.3850129
## Q75
## 2hot_.1_0 0.6134185
## 2hot_.2_0 0.6070288
## 2hot_.3_0 0.6006390
## 2hot_.4_0 0.5942492
## 2hot_.5_0 0.5910543
## 2hot_.6_0 0.5834914
## 2hot_.7_0 0.5826772
## 2hot_.8_0 0.5826772
## 2hot_.9_0 0.5826772
## 2hot_0 0.5826772
##
## $comparison
## NULL
##
## $groups
## size groups
## 2hot_.1_0 0.4880421 a
## 2hot_.2_0 0.4697844 ab
## 2hot_.3_0 0.4581279 bc
## 2hot_.4_0 0.4486385 bcd
## 2hot_.5_0 0.4435669 cde
## 2hot_.6_0 0.4370940 cde
## 2hot_.7_0 0.4340090 cde
## 2hot_.8_0 0.4309716 de
## 2hot_.9_0 0.4243241 de
## 2hot_0 0.4216035 e
##
## attr(,"class")
## [1] "group"sample_size_L <- sample_size[sample_size$infer == 'logistic',]
anova <- aov(size ~ scheme, data = sample_size_L)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## scheme 9 7.1 0.7890 30.5 <2e-16 ***
## Residuals 8990 232.5 0.0259
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, 'scheme', group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 0.02586684 8990 0.4971709 32.34938 0.0239921
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey scheme 10 4.475256 0.05
##
## $means
## size std r Min Max Q25 Q50
## 2hot_.1_0 0.5676679 0.1403551 900 0.2492013 0.9257951 0.4822835 0.5555556
## 2hot_.2_0 0.5190294 0.1535375 900 0.2492013 0.9257951 0.4291339 0.4870127
## 2hot_.3_0 0.5051204 0.1579299 900 0.1565495 0.9257951 0.4122216 0.4680920
## 2hot_.4_0 0.4953881 0.1623954 900 0.1501597 0.9257951 0.3865815 0.4665354
## 2hot_.5_0 0.4944689 0.1602018 900 0.1487805 0.9257951 0.3876812 0.4658537
## 2hot_.6_0 0.4895547 0.1599495 900 0.1565495 0.9257951 0.3769968 0.4527559
## 2hot_.7_0 0.4868893 0.1601321 900 0.1521739 0.9257951 0.3801241 0.4440895
## 2hot_.8_0 0.4820671 0.1614374 900 0.1739130 0.9257951 0.3692471 0.4440895
## 2hot_.9_0 0.4671049 0.1724195 900 0.1501597 0.9257951 0.3557828 0.4440895
## 2hot_0 0.4644182 0.1771996 900 0.1487805 0.9257951 0.3481386 0.4440895
## Q75
## 2hot_.1_0 0.6666667
## 2hot_.2_0 0.6590009
## 2hot_.3_0 0.6371981
## 2hot_.4_0 0.6253230
## 2hot_.5_0 0.6219512
## 2hot_.6_0 0.6223514
## 2hot_.7_0 0.6149871
## 2hot_.8_0 0.6089765
## 2hot_.9_0 0.5891473
## 2hot_0 0.5902439
##
## $comparison
## NULL
##
## $groups
## size groups
## 2hot_.1_0 0.5676679 a
## 2hot_.2_0 0.5190294 b
## 2hot_.3_0 0.5051204 bc
## 2hot_.4_0 0.4953881 bc
## 2hot_.5_0 0.4944689 c
## 2hot_.6_0 0.4895547 cd
## 2hot_.7_0 0.4868893 cde
## 2hot_.8_0 0.4820671 cde
## 2hot_.9_0 0.4671049 de
## 2hot_0 0.4644182 e
##
## attr(,"class")
## [1] "group"sample_size_B <- sample_size[sample_size$infer == 'simple_bayes',]
anova <- aov(size ~ scheme, data = sample_size_B)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## scheme 9 0.47 0.05201 4.195 1.98e-05 ***
## Residuals 8990 111.47 0.01240
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, 'scheme', group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 0.01239903 8990 0.5632614 19.76897 0.0166108
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey scheme 10 4.475256 0.05
##
## $means
## size std r Min Max Q25 Q50
## 2hot_.1_0 0.5835505 0.1019548 900 0.2459016 0.9257951 0.5079589 0.5560041
## 2hot_.2_0 0.5657777 0.1056181 900 0.2903226 0.9257951 0.4891304 0.5354331
## 2hot_.3_0 0.5648956 0.1076096 900 0.3064516 0.9257951 0.4921260 0.5360092
## 2hot_.4_0 0.5621720 0.1097663 900 0.3064516 0.9257951 0.4842520 0.5382130
## 2hot_.5_0 0.5616209 0.1105143 900 0.2741935 0.9257951 0.4881890 0.5359931
## 2hot_.6_0 0.5596227 0.1138995 900 0.2741935 0.9257951 0.4783465 0.5426357
## 2hot_.7_0 0.5593674 0.1147504 900 0.3043478 0.9257951 0.4756882 0.5432190
## 2hot_.8_0 0.5584849 0.1160022 900 0.2741935 0.9257951 0.4763780 0.5413386
## 2hot_.9_0 0.5582443 0.1163769 900 0.2741935 0.9257951 0.4740241 0.5431310
## 2hot_0 0.5588774 0.1160072 900 0.2741935 0.9257951 0.4746377 0.5467850
## Q75
## 2hot_.1_0 0.6291816
## 2hot_.2_0 0.6134185
## 2hot_.3_0 0.6134185
## 2hot_.4_0 0.6134185
## 2hot_.5_0 0.6134185
## 2hot_.6_0 0.6134185
## 2hot_.7_0 0.6134185
## 2hot_.8_0 0.6134185
## 2hot_.9_0 0.6134185
## 2hot_0 0.6134185
##
## $comparison
## NULL
##
## $groups
## size groups
## 2hot_.1_0 0.5835505 a
## 2hot_.2_0 0.5657777 b
## 2hot_.3_0 0.5648956 b
## 2hot_.4_0 0.5621720 b
## 2hot_.5_0 0.5616209 b
## 2hot_.6_0 0.5596227 b
## 2hot_.7_0 0.5593674 b
## 2hot_0 0.5588774 b
## 2hot_.8_0 0.5584849 b
## 2hot_.9_0 0.5582443 b
##
## attr(,"class")
## [1] "group"anova <- aov(size ~ scheme * infer, data = sample_size)
summary(anova)## Df Sum Sq Mean Sq F value Pr(>F)
## scheme 9 8.9 0.993 44.07 < 2e-16 ***
## infer 2 62.6 31.299 1388.72 < 2e-16 ***
## scheme:infer 18 2.2 0.124 5.51 3.34e-13 ***
## Residuals 26970 607.9 0.023
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1print(HSD.test(anova, c('infer', 'scheme'), group=TRUE))## $statistics
## MSerror Df Mean CV MSD
## 0.02253825 26970 0.5020162 29.90491 0.02652897
##
## $parameters
## test name.t ntr StudentizedRange alpha
## Tukey infer:scheme 30 5.30129 0.05
##
## $means
## size std r Min Max
## logistic:2hot_.1_0 0.5676679 0.1403551 900 0.24920128 0.9257951
## logistic:2hot_.2_0 0.5190294 0.1535375 900 0.24920128 0.9257951
## logistic:2hot_.3_0 0.5051204 0.1579299 900 0.15654952 0.9257951
## logistic:2hot_.4_0 0.4953881 0.1623954 900 0.15015974 0.9257951
## logistic:2hot_.5_0 0.4944689 0.1602018 900 0.14878049 0.9257951
## logistic:2hot_.6_0 0.4895547 0.1599495 900 0.15654952 0.9257951
## logistic:2hot_.7_0 0.4868893 0.1601321 900 0.15217391 0.9257951
## logistic:2hot_.8_0 0.4820671 0.1614374 900 0.17391304 0.9257951
## logistic:2hot_.9_0 0.4671049 0.1724195 900 0.15015974 0.9257951
## logistic:2hot_0 0.4644182 0.1771996 900 0.14878049 0.9257951
## once_success:2hot_.1_0 0.4880421 0.1587421 900 0.10335917 0.9257951
## once_success:2hot_.2_0 0.4697844 0.1648237 900 0.10077519 0.9257951
## once_success:2hot_.3_0 0.4581279 0.1691248 900 0.08536585 0.9257951
## once_success:2hot_.4_0 0.4486385 0.1708800 900 0.08010336 0.9257951
## once_success:2hot_.5_0 0.4435669 0.1712949 900 0.07751938 0.9257951
## once_success:2hot_.6_0 0.4370940 0.1725848 900 0.07751938 0.9257951
## once_success:2hot_.7_0 0.4340090 0.1740101 900 0.07235142 0.9257951
## once_success:2hot_.8_0 0.4309716 0.1750216 900 0.06829268 0.9257951
## once_success:2hot_.9_0 0.4243241 0.1770567 900 0.06201550 0.9257951
## once_success:2hot_0 0.4216035 0.1786931 900 0.06201550 0.9257951
## simple_bayes:2hot_.1_0 0.5835505 0.1019548 900 0.24590164 0.9257951
## simple_bayes:2hot_.2_0 0.5657777 0.1056181 900 0.29032258 0.9257951
## simple_bayes:2hot_.3_0 0.5648956 0.1076096 900 0.30645161 0.9257951
## simple_bayes:2hot_.4_0 0.5621720 0.1097663 900 0.30645161 0.9257951
## simple_bayes:2hot_.5_0 0.5616209 0.1105143 900 0.27419355 0.9257951
## simple_bayes:2hot_.6_0 0.5596227 0.1138995 900 0.27419355 0.9257951
## simple_bayes:2hot_.7_0 0.5593674 0.1147504 900 0.30434783 0.9257951
## simple_bayes:2hot_.8_0 0.5584849 0.1160022 900 0.27419355 0.9257951
## simple_bayes:2hot_.9_0 0.5582443 0.1163769 900 0.27419355 0.9257951
## simple_bayes:2hot_0 0.5588774 0.1160072 900 0.27419355 0.9257951
## Q25 Q50 Q75
## logistic:2hot_.1_0 0.4822835 0.5555556 0.6666667
## logistic:2hot_.2_0 0.4291339 0.4870127 0.6590009
## logistic:2hot_.3_0 0.4122216 0.4680920 0.6371981
## logistic:2hot_.4_0 0.3865815 0.4665354 0.6253230
## logistic:2hot_.5_0 0.3876812 0.4658537 0.6219512
## logistic:2hot_.6_0 0.3769968 0.4527559 0.6223514
## logistic:2hot_.7_0 0.3801241 0.4440895 0.6149871
## logistic:2hot_.8_0 0.3692471 0.4440895 0.6089765
## logistic:2hot_.9_0 0.3557828 0.4440895 0.5891473
## logistic:2hot_0 0.3481386 0.4440895 0.5902439
## once_success:2hot_.1_0 0.3731884 0.4520241 0.6134185
## once_success:2hot_.2_0 0.3562992 0.4440895 0.6070288
## once_success:2hot_.3_0 0.3322684 0.4237726 0.6006390
## once_success:2hot_.4_0 0.3226837 0.4121406 0.5942492
## once_success:2hot_.5_0 0.3180212 0.4089457 0.5910543
## once_success:2hot_.6_0 0.3041868 0.4057508 0.5834914
## once_success:2hot_.7_0 0.2966889 0.4057508 0.5826772
## once_success:2hot_.8_0 0.2962963 0.3984610 0.5826772
## once_success:2hot_.9_0 0.2907348 0.3851894 0.5826772
## once_success:2hot_0 0.2834646 0.3850129 0.5826772
## simple_bayes:2hot_.1_0 0.5079589 0.5560041 0.6291816
## simple_bayes:2hot_.2_0 0.4891304 0.5354331 0.6134185
## simple_bayes:2hot_.3_0 0.4921260 0.5360092 0.6134185
## simple_bayes:2hot_.4_0 0.4842520 0.5382130 0.6134185
## simple_bayes:2hot_.5_0 0.4881890 0.5359931 0.6134185
## simple_bayes:2hot_.6_0 0.4783465 0.5426357 0.6134185
## simple_bayes:2hot_.7_0 0.4756882 0.5432190 0.6134185
## simple_bayes:2hot_.8_0 0.4763780 0.5413386 0.6134185
## simple_bayes:2hot_.9_0 0.4740241 0.5431310 0.6134185
## simple_bayes:2hot_0 0.4746377 0.5467850 0.6134185
##
## $comparison
## NULL
##
## $groups
## size groups
## simple_bayes:2hot_.1_0 0.5835505 a
## logistic:2hot_.1_0 0.5676679 a
## simple_bayes:2hot_.2_0 0.5657777 a
## simple_bayes:2hot_.3_0 0.5648956 a
## simple_bayes:2hot_.4_0 0.5621720 a
## simple_bayes:2hot_.5_0 0.5616209 a
## simple_bayes:2hot_.6_0 0.5596227 a
## simple_bayes:2hot_.7_0 0.5593674 a
## simple_bayes:2hot_0 0.5588774 a
## simple_bayes:2hot_.8_0 0.5584849 a
## simple_bayes:2hot_.9_0 0.5582443 a
## logistic:2hot_.2_0 0.5190294 b
## logistic:2hot_.3_0 0.5051204 bc
## logistic:2hot_.4_0 0.4953881 bcd
## logistic:2hot_.5_0 0.4944689 bcd
## logistic:2hot_.6_0 0.4895547 cde
## once_success:2hot_.1_0 0.4880421 cde
## logistic:2hot_.7_0 0.4868893 cde
## logistic:2hot_.8_0 0.4820671 cdef
## once_success:2hot_.2_0 0.4697844 defg
## logistic:2hot_.9_0 0.4671049 efg
## logistic:2hot_0 0.4644182 efg
## once_success:2hot_.3_0 0.4581279 fgh
## once_success:2hot_.4_0 0.4486385 ghi
## once_success:2hot_.5_0 0.4435669 ghij
## once_success:2hot_.6_0 0.4370940 hij
## once_success:2hot_.7_0 0.4340090 hij
## once_success:2hot_.8_0 0.4309716 ij
## once_success:2hot_.9_0 0.4243241 ij
## once_success:2hot_0 0.4216035 j
##
## attr(,"class")
## [1] "group"