/
icse14-mut-beta.R
122 lines (90 loc) · 3.84 KB
/
icse14-mut-beta.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
#
# icse14-mut-beta.R, 21 Sep 18
#
# Data from:
# Code Coverage for Suite Evaluation by Developers
# Rahul Gopinath and Carlos Jensen and Alex Groce
#
# Example from:
# Evidence-based Software Engineering: based on the publicly available data
# Derek M. Jones
#
# TAG coverage_testing testing_mutation
source("ESEUR_config.r")
library("betareg")
library("boot")
# plot_layout(2, 2)
line_col=rainbow(3)
load(paste0(ESEUR_dir, "reliability/o.db.rda"))
cov_info=subset(o.db, loc > 0)
cov_info$log_loc=log(cov_info$loc)
cov_info$line_cov=cov_info$cobertura.line.num/cov_info$cobertura.line.total
cov_info$branch_cov=cov_info$cobertura.branch.num/cov_info$cobertura.branch.total
cov_info$block_cov=cov_info$nemma.block.num/cov_info$nemma.block.total
cov_info$path_cov=cov_info$mockit.path.num/cov_info$mockit.path.total
cov_info$mut_cov=cov_info$pit.mutation.num/cov_info$pit.mutation.total
max_loc=max(cov_info$log_loc)
boot_reg=function(data, indices)
{
cov_data = data[indices, ] # select obs. in bootstrap sample
b_mod=betareg(mut_cov ~ I(path_cov^0.5), data=cov_data)
# A vector must be returned, i.e., no dataframes
return(predict(b_mod, newdata=data.frame(path_cov=x_vals)))
}
mut_coverage=function(x_measure, y_measure, x_str, y_str, legend_pos)
{
applic=(x_measure > 0) & (y_measure > 0) & (y_measure < 1) & !(is.na(x_measure) | is.na(y_measure))
x_measure=x_measure[applic]
y_measure=y_measure[applic]
# print(c(x_measure[135:145], y_measure[135:145]))
log_loc=subset(cov_info, applic)$log_loc
plot(x_measure, y_measure, type="n",
xlab=x_str, ylab=paste0(y_str, "\n"))
pal_col=rainbow(1+max_loc)
symbols(x_measure, y_measure, circles=log_loc,
inches=1/15, add=T, fg=pal_col[log_loc+1])
legend(legend_pos,
c(expression(LOC %~~% 10^1), expression(LOC %~~% 10^3), expression(LOC %~~% 10^5)),
pch=21, bty='n',
pt.cex=c(0.5, 1.0, 1.7),
col=c(pal_col[1+2], pal_col[1+6], pal_col[1+max_loc]))
cov_mod=betareg(y_measure ~ I(x_measure^0.5))
pred=predict(cov_mod, newdata=data.frame(x_measure=x_vals))
lines(x_vals, pred, col=line_col[1])
# q_pred=predict(cov_mod, newdata=data.frame(x_measure=x_vals),
# type="quantile", at=c(0.025, 0.975))
# lines(x_vals, q_pred[, 1], col=line_col[1])
# lines(x_vals, q_pred[, 2], col=line_col[1])
cov_gmod=glm(y_measure ~ I(x_measure^0.5))
g_pred=predict(cov_gmod, newdata=data.frame(x_measure=x_vals))
lines(x_vals, g_pred, col=line_col[2])
# lines(loess.smooth(x_measure, y_measure, span=0.3), col="black")
return(cov_mod)
}
x_vals=seq(0, 1, by=0.01)
# mut_coverage(cov_info$line_cov, cov_info$mut_cov,
# "Statement coverage", "Percentage mutants killed", "bottomright")
mod=mut_coverage(cov_info$path_cov, cov_info$mut_cov,
"Path coverage", "Percentage mutants killed", "bottomright")
# summary(mod)
pm_info=subset(cov_info, (path_cov > 0) & (mut_cov > 0) & (mut_cov < 1) &
!(is.na(path_cov) | is.na(mut_cov)))
cov_boot=boot(pm_info, boot_reg, R = 4999)
ci=apply(cov_boot$t, 2, function(X) quantile(X, c(0.025, 0.975)))
lines(x_vals, ci[1, ], col=line_col[3])
lines(x_vals, ci[2, ], col=line_col[3])
# clean_cov=subset(cov_info, (path_cov > 0) & (mut_cov > 0) & (mut_cov < 1) & !(is.na(path_cov) | is.na(mut_cov)))
# cov_nls_mod=nls(mut_cov ~ a+b*path_cov^c, data=clean_cov,
# start=list(a=0, b=1, c=0.5))
# cov_bmod=betareg(mut_cov ~ I(line_cov^0.5), data=clean_cov)
# pred=predict(cov_nls_mod, newdata=data.frame(path_cov=x_vals))
# lines(x_vals, pred, col="green")
# mut_coverage(cov_info$line_cov, cov_info$path_cov,
# "Statement coverage", "Path coverage", "topleft")
# mut_coverage(cov_info$line_cov, cov_info$block_cov,
# "Statement coverage", "Block coverage", "topleft")
# mut_coverage(cov_info$line_cov, cov_info$branch_cov,
# "Statement coverage", "Branch coverage", "topleft")
# plot(cov_info$loc, cov_info$line_cov, log="x",
# xlab="Lines of code", ylab="Statement coverage")
# cov_mod=glm(log_loc ~ line_cov, data=cov_info)