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
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
|
# Copyright (C) 2010-2014 Johannes Ranke
# Contact: jranke@uni-bremen.de
# This file is part of the R package mkin
# mkin is free software: you can redistribute it and/or modify it under the
# terms of the GNU General Public License as published by the Free Software
# Foundation, either version 3 of the License, or (at your option) any later
# version.
# This program is distributed in the hope that it will be useful, but WITHOUT
# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
# FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
# details.
# You should have received a copy of the GNU General Public License along with
# this program. If not, see <http://www.gnu.org/licenses/>
# Test SFO model to a relative tolerance of 1% # {{{
test.FOCUS_2006_SFO <- function()
{
SFO.1 <- mkinmod(parent = list(type = "SFO"))
SFO.2 <- mkinmod(parent = list(type = "SFO"), use_of_ff = "max")
fit.A.SFO.1 <- mkinfit(SFO.1, FOCUS_2006_A, quiet=TRUE)
fit.A.SFO.2 <- mkinfit(SFO.2, FOCUS_2006_A, quiet=TRUE)
median.A.SFO <- as.numeric(lapply(subset(FOCUS_2006_SFO_ref_A_to_F, dataset == "A",
c(M0, k, DT50, DT90)), "median"))
fit.A.SFO.1.r <- as.numeric(c(fit.A.SFO.1$bparms.optim, endpoints(fit.A.SFO.1)$distimes))
dev.A.SFO.1 <- abs(round(100 * ((median.A.SFO - fit.A.SFO.1.r)/median.A.SFO), digits=1))
checkIdentical(dev.A.SFO.1 < 1, rep(TRUE, length(dev.A.SFO.1)))
fit.A.SFO.2.r <- as.numeric(c(fit.A.SFO.2$bparms.optim, endpoints(fit.A.SFO.2)$distimes))
dev.A.SFO.2 <- abs(round(100 * ((median.A.SFO - fit.A.SFO.2.r)/median.A.SFO), digits=1))
checkIdentical(dev.A.SFO.2 < 1, rep(TRUE, length(dev.A.SFO.2)))
fit.C.SFO.1 <- mkinfit(SFO.1, FOCUS_2006_C, quiet=TRUE)
fit.C.SFO.2 <- mkinfit(SFO.2, FOCUS_2006_C, quiet=TRUE)
median.C.SFO <- as.numeric(lapply(subset(FOCUS_2006_SFO_ref_A_to_F, dataset == "C",
c(M0, k, DT50, DT90)), "median"))
fit.C.SFO.1.r <- as.numeric(c(fit.C.SFO.1$bparms.optim, endpoints(fit.C.SFO.1)$distimes))
dev.C.SFO.1 <- abs(round(100 * ((median.C.SFO - fit.C.SFO.1.r)/median.C.SFO), digits=1))
checkIdentical(dev.C.SFO.1 < 1, rep(TRUE, length(dev.C.SFO.1)))
fit.C.SFO.2.r <- as.numeric(c(fit.C.SFO.2$bparms.optim, endpoints(fit.C.SFO.2)$distimes))
dev.C.SFO.2 <- abs(round(100 * ((median.C.SFO - fit.C.SFO.2.r)/median.C.SFO), digits=1))
checkIdentical(dev.C.SFO.2 < 1, rep(TRUE, length(dev.C.SFO.2)))
} # }}}
# Test FOMC model to a relative tolerance of 1% {{{
# See kinfit vignette for a discussion of FOMC fits to FOCUS_2006_A
# In this case, only M0, DT50 and DT90 are checked
test.FOCUS_2006_FOMC <- function()
{
FOMC <- mkinmod(parent = list(type = "FOMC"))
# FOCUS_2006_A (compare kinfit vignette for discussion)
fit.A.FOMC <- mkinfit(FOMC, FOCUS_2006_A, quiet=TRUE)
median.A.FOMC <- as.numeric(lapply(subset(FOCUS_2006_FOMC_ref_A_to_F, dataset == "A",
c(M0, alpha, beta, DT50, DT90)), "median"))
fit.A.FOMC.r <- as.numeric(c(fit.A.FOMC$bparms.optim, endpoints(fit.A.FOMC)$distimes[c("DT50", "DT90")]))
dev.A.FOMC <- abs(round(100 * ((median.A.FOMC - fit.A.FOMC.r)/median.A.FOMC), digits=1))
dev.A.FOMC <- dev.A.FOMC[c(1, 4, 5)]
checkIdentical(dev.A.FOMC < 1, rep(TRUE, length(dev.A.FOMC)))
# FOCUS_2006_B
fit.B.FOMC <- mkinfit(FOMC, FOCUS_2006_B, quiet=TRUE)
median.B.FOMC <- as.numeric(lapply(subset(FOCUS_2006_FOMC_ref_A_to_F, dataset == "B",
c(M0, alpha, beta, DT50, DT90)), "median"))
fit.B.FOMC.r <- as.numeric(c(fit.B.FOMC$bparms.optim, endpoints(fit.B.FOMC)$distimes[c("DT50", "DT90")]))
dev.B.FOMC <- abs(round(100 * ((median.B.FOMC - fit.B.FOMC.r)/median.B.FOMC), digits=1))
dev.B.FOMC <- dev.B.FOMC[c(1, 4, 5)]
checkIdentical(dev.B.FOMC < 1, rep(TRUE, length(dev.B.FOMC)))
# FOCUS_2006_C
fit.C.FOMC <- mkinfit(FOMC, FOCUS_2006_C, quiet=TRUE)
median.C.FOMC <- as.numeric(lapply(subset(FOCUS_2006_FOMC_ref_A_to_F, dataset == "C",
c(M0, alpha, beta, DT50, DT90)), "median"))
fit.C.FOMC.r <- as.numeric(c(fit.C.FOMC$bparms.optim, endpoints(fit.C.FOMC)$distimes[c("DT50", "DT90")]))
dev.C.FOMC <- abs(round(100 * ((median.C.FOMC - fit.C.FOMC.r)/median.C.FOMC), digits=1))
dev.C.FOMC <- dev.C.FOMC[c(1, 4, 5)]
checkIdentical(dev.C.FOMC < 1, rep(TRUE, length(dev.C.FOMC)))
} # }}}
# Test DFOP model, tolerance of 1% with the exception of f parameter for A {{{
test.FOCUS_2006_DFOP <- function()
{
DFOP <- mkinmod(parent = list(type = "DFOP"))
# FOCUS_2006_A
# Results were too much dependent on algorithm, as this dataset
# is pretty much SFO. "Port" gave a lower deviance, but deviated from the
# median of FOCUS_2006 solutions
# FOCUS_2006_B
fit.B.DFOP <- mkinfit(DFOP, FOCUS_2006_B, quiet=TRUE)
median.B.DFOP <- as.numeric(lapply(subset(FOCUS_2006_DFOP_ref_A_to_B, dataset == "B",
c(M0, k1, k2, f, DT50, DT90)), "median"))
fit.B.DFOP.r <- as.numeric(c(fit.B.DFOP$bparms.optim, endpoints(fit.B.DFOP)$distimes[c("DT50", "DT90")]))
dev.B.DFOP <- abs(round(100 * ((median.B.DFOP - fit.B.DFOP.r)/median.B.DFOP), digits=1))
# about 0.6% deviation for parameter f, the others are <= 0.1%
checkIdentical(dev.B.DFOP < 1, rep(TRUE, length(dev.B.DFOP)))
} # }}}
# Test HS model to a relative tolerance of 1% excluding Mathematica values {{{
# as they are unreliable
test.FOCUS_2006_HS <- function()
{
HS <- mkinmod(parent = list(type = "HS"))
# FOCUS_2006_A
fit.A.HS <- mkinfit(HS, FOCUS_2006_A, quiet=TRUE)
median.A.HS <- as.numeric(lapply(subset(FOCUS_2006_HS_ref_A_to_F, dataset == "A",
c(M0, k1, k2, tb, DT50, DT90)), "median"))
fit.A.HS.r <- as.numeric(c(fit.A.HS$bparms.optim, endpoints(fit.A.HS)$distimes[c("DT50", "DT90")]))
dev.A.HS <- abs(round(100 * ((median.A.HS - fit.A.HS.r)/median.A.HS), digits=1))
# about 6.7% deviation for parameter f, the others are < 0.1%
checkIdentical(dev.A.HS < 1, rep(TRUE, length(dev.A.HS)))
# FOCUS_2006_B
fit.B.HS <- mkinfit(HS, FOCUS_2006_B, quiet=TRUE)
median.B.HS <- as.numeric(lapply(subset(FOCUS_2006_HS_ref_A_to_F, dataset == "B",
c(M0, k1, k2, tb, DT50, DT90)), "median"))
fit.B.HS.r <- as.numeric(c(fit.B.HS$bparms.optim, endpoints(fit.B.HS)$distimes[c("DT50", "DT90")]))
dev.B.HS <- abs(round(100 * ((median.B.HS - fit.B.HS.r)/median.B.HS), digits=1))
# < 10% deviation for M0, k1, DT50 and DT90, others are problematic
dev.B.HS <- dev.B.HS[c(1, 2, 5, 6)]
checkIdentical(dev.B.HS < 10, rep(TRUE, length(dev.B.HS)))
# FOCUS_2006_C
fit.C.HS <- mkinfit(HS, FOCUS_2006_C, quiet=TRUE)
median.C.HS <- as.numeric(lapply(subset(FOCUS_2006_HS_ref_A_to_F, dataset == "C",
c(M0, k1, k2, tb, DT50, DT90)), "median"))
fit.C.HS.r <- as.numeric(c(fit.C.HS$bparms.optim, endpoints(fit.C.HS)$distimes[c("DT50", "DT90")]))
dev.C.HS <- abs(round(100 * ((median.C.HS - fit.C.HS.r)/median.C.HS), digits=1))
# deviation <= 0.1%
checkIdentical(dev.C.HS < 1, rep(TRUE, length(dev.C.HS)))
} # }}}
# Test SFORB model against DFOP solutions to a relative tolerance of 1% # {{{
test.FOCUS_2006_SFORB <- function()
{
SFORB <- mkinmod(parent = list(type = "SFORB"))
# FOCUS_2006_A
# Again it does not make a lot of sense to use a SFO dataset for this
# FOCUS_2006_B
fit.B.SFORB.1 <- mkinfit(SFORB, FOCUS_2006_B, quiet=TRUE)
fit.B.SFORB.2 <- mkinfit(SFORB, FOCUS_2006_B, solution_type = "deSolve", quiet=TRUE)
median.B.SFORB <- as.numeric(lapply(subset(FOCUS_2006_DFOP_ref_A_to_B, dataset == "B",
c(M0, k1, k2, DT50, DT90)), "median"))
fit.B.SFORB.1.r <- as.numeric(c(
parent_0 = fit.B.SFORB.1$bparms.optim[[1]],
k1 = endpoints(fit.B.SFORB.1)$SFORB[[1]],
k2 = endpoints(fit.B.SFORB.1)$SFORB[[2]],
endpoints(fit.B.SFORB.1)$distimes[c("DT50", "DT90")]))
dev.B.SFORB.1 <- abs(round(100 * ((median.B.SFORB - fit.B.SFORB.1.r)/median.B.SFORB), digits=1))
checkIdentical(dev.B.SFORB.1 < 1, rep(TRUE, length(dev.B.SFORB.1)))
fit.B.SFORB.2.r <- as.numeric(c(
parent_0 = fit.B.SFORB.2$bparms.optim[[1]],
k1 = endpoints(fit.B.SFORB.2)$SFORB[[1]],
k2 = endpoints(fit.B.SFORB.2)$SFORB[[2]],
endpoints(fit.B.SFORB.2)$distimes[c("DT50", "DT90")]))
dev.B.SFORB.2 <- abs(round(100 * ((median.B.SFORB - fit.B.SFORB.2.r)/median.B.SFORB), digits=1))
checkIdentical(dev.B.SFORB.2 < 1, rep(TRUE, length(dev.B.SFORB.2)))
} # }}}
# Test eigenvalue based fit to Schaefer 2007 data against solution from conference paper {{{
test.mkinfit.schaefer07_complex_example <- function()
{
schaefer07_complex_model <- mkinmod(
parent = list(type = "SFO", to = c("A1", "B1", "C1"), sink = FALSE),
A1 = list(type = "SFO", to = "A2"),
B1 = list(type = "SFO"),
C1 = list(type = "SFO"),
A2 = list(type = "SFO"), use_of_ff = "max")
# If we use the default algorithm 'Marq' we need to give a good starting
# estimate for k_A2 in order to find the solution published by Schaefer et al.
fit <- mkinfit(schaefer07_complex_model, method.modFit = "Port",
mkin_wide_to_long(schaefer07_complex_case, time = "time"))
s <- summary(fit)
r <- schaefer07_complex_results
attach(as.list(fit$bparms.optim))
r$mkin <- c(
k_parent,
s$distimes["parent", "DT50"],
s$ff["parent_A1"],
k_A1,
s$distimes["A1", "DT50"],
s$ff["parent_B1"],
k_B1,
s$distimes["B1", "DT50"],
s$ff["parent_C1"],
k_C1,
s$distimes["C1", "DT50"],
s$ff["A1_A2"],
k_A2,
s$distimes["A2", "DT50"])
r$means <- (r$KinGUI + r$ModelMaker)/2
r$mkin.deviation <- abs(round(100 * ((r$mkin - r$means)/r$means), digits=1))
checkIdentical(r$mkin.deviation < 10, rep(TRUE, 14))
} # }}}
# vim: set foldmethod=marker ts=2 sw=2 expandtab:
|
