-
Notifications
You must be signed in to change notification settings - Fork 0
/
reprogramming.py
174 lines (152 loc) · 7.17 KB
/
reprogramming.py
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
# Copyright or © or Copr. Loïc Paulevé (2023)
#
# loic.pauleve@cnrs.fr
#
# This software is governed by the CeCILL license under French law and
# abiding by the rules of distribution of free software. You can use,
# modify and/ or redistribute the software under the terms of the CeCILL
# license as circulated by CEA, CNRS and INRIA at the following URL
# "http://www.cecill.info".
#
# As a counterpart to the access to the source code and rights to copy,
# modify and redistribute granted by the license, users are provided only
# with a limited warranty and the software's author, the holder of the
# economic rights, and the successive licensors have only limited
# liability.
#
# In this respect, the user's attention is drawn to the risks associated
# with loading, using, modifying and/or developing or reproducing the
# software by the user in light of its specific status of free software,
# that may mean that it is complicated to manipulate, and that also
# therefore means that it is reserved for developers and experienced
# professionals having in-depth computer knowledge. Users are therefore
# encouraged to load and test the software's suitability as regards their
# requirements in conditions enabling the security of their systems and/or
# data to be ensured and, more generally, to use and operate it in the
# same conditions as regards security.
#
# The fact that you are presently reading this means that you have had
# knowledge of the CeCILL license and that you accept its terms.
#
import bonesis
def marker_reprogramming_fixpoints(f, M, k, at_least_one=True, **some_opts):
bo = bonesis.BoNesis(f)
control = bo.Some(max_size=k, **some_opts)
with bo.mutant(control):
if at_least_one:
bo.fixed(~bo.obs(M))
bo.all_fixpoints(bo.obs(M))
return control.assignments()
def source_marker_reprogramming_fixpoints(f, z, M, k, at_least_one=True,
**some_opts):
bo = bonesis.BoNesis(f)
control = bo.Some(max_size=k, **some_opts)
with bo.mutant(control):
if at_least_one:
~bo.obs(z) >= bo.fixed(~bo.obs(M))
~bo.obs(z) >> "fixpoints" ^ {bo.obs(M)}
return control.assignments()
def trapspace_reprogramming(f, M, k, algorithm="cegar", **some_opts):
"""
Marker/Trapspace reprogramming of BN domain `f`.
Returns mutations which ensure that all the minimal trap spaces match with
the given marker `M`.
"""
if algorithm == "cegar":
meth = _trapspace_reprogramming_cegar
else:
meth = _trapspace_reprogramming_complementary
return meth(f, M, k, **some_opts)
marker_reprogramming = trapspace_reprogramming
def _trapspace_reprogramming_complementary(f, M, k, **some_opts):
bo = bonesis.BoNesis(f)
bad_control = bo.Some(max_size=k, **some_opts)
with bo.mutant(bad_control):
x = bo.cfg()
bo.in_attractor(x) != bo.obs(M)
return bad_control.complementary_assignments()
def source_marker_reprogramming(f, z, M, k, **some_opts):
bo = bonesis.BoNesis(f)
bad_control = bo.Some(max_size=k, **some_opts)
with bo.mutant(bad_control):
x = bo.cfg()
~bo.obs(z) >= bo.in_attractor(x) != bo.obs(M)
return bad_control.complementary_assignments()
###
### CEGAR-based implementations
###
from bonesis.views import SomeView
def _trapspace_reprogramming_cegar(dom, M, k, **some_opts):
data = {"M": M}
bo = bonesis.BoNesis(dom, data)
P = bo.Some(max_size=k, **some_opts)
with bo.mutant(P):
# there exists at least one minimal trap spaces matching M
bo.fixed(bo.obs("M"))
class BoCegar(SomeView):
single_shot = False
project = False
ice = 0
def __iter__(self):
self.configure()
return self
def __next__(self):
while True:
with self.control.solve(yield_=True) as candidates:
found_P = False
for candidate in candidates:
found_P = True
P = self.parse_model(candidate)
## check candidate
boc = bonesis.BoNesis(dom, data)
with boc.mutant(P):
x = boc.cfg()
boc.in_attractor(x)
x != boc.obs("M")
view = x.assignments(limit=1)
view.configure()
if (ce := next(iter(view.control.solve(yield_=True)), None)) is not None:
self.ice += 1
x = [a for a in ce.symbols(shown=True)
if a.name == "cfg" and a.arguments[0].string == "__cfg0"]
else:
p = candidate.symbols(shown=True)
break
if not found_P:
raise StopIteration
if ce is None:
inject = f":- {','.join(map(str, p))}."
self.control.add("skip", [], inject)
self.control.ground([("skip", [])])
return P
ns = f"_ce{self.ice}_"
cets = f"{ns}ts"
fixts = f"{ns}fix"
inject = [f"mcfg({cets},{a.arguments[1]},{a.arguments[2]})." for a in x]
inject += [
# compute trap space of counter example
f"mcfg({cets},N,V) :- {ns}eval({cets},N,V).",
f"clamped({cets},N,V) :- mutant(_,N,V).",
# choose sub-space in it
f"1 {{mcfg({fixts},N,V):mcfg({cets},N,V)}} :- mcfg({cets},N,_).",
# saturate it
f"mcfg({fixts},N,V) :- {ns}eval({fixts},N,V).",
f"clamped({fixts},N,V) :- clamped({cets},N,V).",
f"{ns}eval(X,N,C,-1) :- clause(N,C,L,-V), mcfg(X,L,V), not clamped(X,N,_).",
f"{ns}eval(X,N,C,1) :- mcfg(X,L,V): clause(N,C,L,V); clause(N,C,_,_), mcfg(X,_,_), not clamped(X,N,_).",
f"{ns}eval(X,N,1) :- {ns}eval(X,N,C,1), clause(N,C,_,_).",
f"{ns}eval(X,N,-1) :- {ns}eval(X,N,C,-1): clause(N,C,_,_); clause(N,_,_,_), mcfg(X,_,_).",
f"{ns}eval(X,N,V) :- clamped(X,N,V).",
f"{ns}eval(X,N,V) :- constant(N,V), mcfg(X,_,_), not clamped(X,N,_).",
f"{ns}eval(X,N,V) :- {ns}evalbdd(X,N,V), node(N), not clamped(X,N,_).",
f"{ns}evalbdd(X,V,V) :- mcfg(X,_,_), V=(-1;1).",
f"{ns}evalbdd(X,B,V) :- bdd(B,N,_,HI), mcfg(X,N,1), {ns}evalbdd(X,HI,V).",
f"{ns}evalbdd(X,B,V) :- bdd(B,N,LO,_), mcfg(X,N,-1), {ns}evalbdd(X,LO,V).",
f"{ns}evalbdd(X,B,V) :- mcfg(X,_,_), bdd(B,V).",
]
# TS(y) must match with M
inject.append(f":- obs(\"M\",N,V), mcfg({fixts},N,-V).")
inject = "\n".join(inject)
self.control.add(f"cegar{ns}", [], inject)
self.control.ground([(f"cegar{ns}", [])])
return BoCegar(P, bo, solutions="subset-minimal")