Merge branch 'development' into 'master'

Merging Development into Master



See merge request !2

CHANGELOG.md

@@ -3,6 +3,22 @@ All notable changes to this project will be documented in this file.
 This project adheres to [Semantic Versioning](http://semver.org/).
 
 ## [Unreleased][unreleased]
+### Added
+- 
+- 
+
+### Changed
+- 
+- 
+
+## [v0.3.0] - 2016-01-10
+### Added
+- Case study Tic Tac Toe
+- LearnLib lib for Tic Tac Toe
+
+### Changed
+- License
+- Name: new name is Alnos
 
 ## [v0.2.0] - 2015-10-27
 ### Added
@@ -24,5 +40,7 @@ This project adheres to [Semantic Versioning](http://semver.org/).
 - Test classes
 - Simple Examples
 
-[unreleased]: https://gitlab.science.ru.nl/mvolpato/active-learning-nondeterministic-systems/compare/v0.1.0...HEAD
+[unreleased]: https://gitlab.science.ru.nl/mvolpato/active-learning-nondeterministic-systems/compare/v0.3.0...HEAD
 [v0.1.0]: https://gitlab.science.ru.nl/mvolpato/active-learning-nondeterministic-systems/compare/f7f05033cf5e002a45a67632e60b311892ca0850...v0.1.0
+[v0.2.0]: https://gitlab.science.ru.nl/mvolpato/active-learning-nondeterministic-systems/compare/v0.1.0...v0.2.0
+[v0.3.0]: https://gitlab.science.ru.nl/mvolpato/active-learning-nondeterministic-systems/compare/v0.2.0...v0.3.0
\ No newline at end of file

README.md

 ## Synopsis
 
-The active-learning-nondeterministic-systems is an implementation of an
+**Alnos** is an implementation of an
 adaptation of
 [L*](http://www.cs.berkeley.edu/~dawnsong/teaching/s10/papers/angluin87.pdf) to
 nondeterministic systems. The code is based on these scientific papers:
@@ -14,7 +14,6 @@ simulation, or model checking.
 
 The project is coded in Python3 and tested using Python3.4.
 
-
 ## Included Libraries
 
 [NumPy](https://github.com/numpy/numpy)
@@ -31,35 +30,10 @@ This code exists as a support implementation to the papers mentioned previously.
 
 Clone the repository. Then you can modify any file in [examples](examples/),
 or create your own.
-If you are using a real system under testing (an actual running black box
-  software), you need to write your own adapters to connect it to the
-  learning tool. The adapters should inherit from `AbstractTeacher` in
-  [baseteacher.py](teachers/baseteacher.py) and from `AbstractOracle` in
-  [baseoracle.py](teachers/baseoracle.py).
-
-  In particular, the adapters should implement the abstract methods in
-  `AbstractTeacher` and `AbstractOracle`. Those method are used by the
-  learner to ask so called **output** and **observation** queries.
-
-  Then you can start learning your system:
-
-```python
-teacher = YourOwnAdapterTeacher()
-oracle = YourOwnAdapterOracle()
 
-underModel, overModel = LearningAlgorithm(teacher, oracle, maxLoops=10,
-                            tablePreciseness = 10000, modelPreciseness = 0.1,
-			    tester=tester)
-```
-
-where `underModel` and `overModel` are the under and over approximations
-of your system, respectively, `maxLoops` is the limit of learning loops
-when the learned models are not changing any more, `tablePreciseness` and
-`modelPreciseness` are the levels of preciseness you would like to reach
-before stopping. Tester is a testing algorithm.
-
-The learning process stops when either the learned model does not change for
-`maxLoops` loops, or when both the preciseness levels are met.
+[Tic Tac Toe](examples/tictactoe/) uses a real *black box* system under learning.
+There is also an example for learning a model of it using another learning
+tool: [LearnLib](https://github.com/LearnLib/learnlib).
 
 ## Contributors
 
@@ -68,4 +42,6 @@ checking [my contact details](https://gitlab.science.ru.nl/u/mvolpato).
 
 ## License
 
-See [LICENSE](./LICENSE)
+[This license](./LICENSE) applies to most of the files. Some files may have a different license. If so, the license can be found at the top of the source code.
+
+If no license is found at the top of the source, [this license](./LICENSE) is applied.

examples/figure2-5_thesis.py

+# Copyright (c) 2015 Michele Volpato
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+# THE SOFTWARE.
+
+# This file is used to learn the IOTS depicted in Figure 2.5 of my PhD thesis.
+
+import random
+
+seed = output = random.sample(range(99999999), 1)[0]
+print(seed)
+random.seed(81077353)   # 81077353
+
+import os, inspect, sys
+# Include project dir in path
+currentdir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
+parentdir = os.path.dirname(currentdir)
+sys.path.append(parentdir)
+from learning.learning import LearningAlgorithm
+from teachers.ltsteachers import InputOutputTeacher
+from systems.implementations import InputOutputLTS
+from teachers.ltsoracles import InputOutputPowerOracle
+import logging
+import helpers.bisimulation as bi
+from testing.randomtesting import RandomTester
+from systems.iopurpose import InputPurpose, OutputPurpose
+import helpers.graphhelper as gh
+
+
+logging.basicConfig(level=logging.DEBUG)
+logger = logging.getLogger(__name__)
+
+inputs = set(['b'])
+outputs = set(['t','c'])
+quiescence = 'd'
+
+I1=InputOutputLTS(8, inputs, outputs, quiescence)
+I1.addTransition(0,'b',1)
+I1.addTransition(0,'b',2)
+I1.addTransition(1,'t',3)
+I1.addTransition(1,'c',3)
+I1.addTransition(1,'b',6)
+I1.addTransition(2,'b',4)
+I1.addTransition(3,'t',0)
+I1.addTransition(3,'c',0)
+I1.addTransition(3,'b',6)
+I1.addTransition(4,'c',5)
+I1.addTransition(4,'b',6)
+I1.addTransition(5,'b',6)
+I1.addTransition(5,'t',0)
+I1.addTransition(5,'c',0)
+
+# Chaos
+I1.addTransition(6,'b',6)
+I1.addTransition(6,'t',6)
+I1.addTransition(6,'c',6)
+I1.addTransition(6,'d',7)
+I1.addTransition(7,'b',6)
+
+
+I1.makeInputEnabled()
+
+T1 = InputOutputTeacher(I1)
+O1 = InputOutputPowerOracle(I1)
+
+outputExpert = OutputPurpose(set(['t','c', quiescence]))
+inputExpert = InputPurpose(set(['b']))
+
+tester = RandomTester(T1, 10000, 50)
+
+currentdir = os.path.dirname(os.path.abspath(
+                inspect.getfile(inspect.currentframe())))
+
+path = os.path.join(currentdir, "dotFiles")
+gh.createDOTFile(I1, path + "figure2-5", "pdf")
+
+print("Starting learning...")
+
+# change printPath=None to printPath=path for dot files
+L2 = LearningAlgorithm(T1, O1, printPath=path, maxLoops=4,
+    tablePreciseness=10000, logger=logger, tester=tester, outputPurpose=outputExpert,
+    inputPurpose=inputExpert)
+minus, plus = L2.run()
+
+print("Models learned. Check language equivalence...")
+
+print("hMinus bisimilar to target: " + str(bi.bisimilar(I1,minus)))
+print("hPlus bisimilar to target: " + str(bi.bisimilar(I1,plus)))

examples/iots.py

+# Copyright (c) 2015 Michele Volpato
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+# THE SOFTWARE.
+
 import os,sys,inspect
 # Include project dir in path
 currentdir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))

examples/learn.py

+# Copyright (c) 2015 Michele Volpato
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+# THE SOFTWARE.
+
+import random
+
+seed = output = random.sample(range(99999999), 1)[0]
+print(seed)
+random.seed(81077353)   # 81077353
+
 import os, inspect, sys
 # Include project dir in path
 currentdir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
@@ -10,8 +36,10 @@ from teachers.ltsoracles import InputOutputPowerOracle
 import logging
 import helpers.bisimulation as bi
 from testing.randomtesting import RandomTester
+from systems.iopurpose import InputPurpose, OutputPurpose
 
-logging.basicConfig(level=logging.INFO)
+
+logging.basicConfig(level=logging.DEBUG)
 logger = logging.getLogger(__name__)
 
 inputs = set(['a','b'])
@@ -48,7 +76,10 @@ I1.makeInputEnabled()
 T1 = InputOutputTeacher(I1)
 O1 = InputOutputPowerOracle(I1)
 
-tester = RandomTester(T1, 10000, 20)
+outputExpert = OutputPurpose(set(['x','y', quiescence]))
+inputExpert = InputPurpose(set(['a','b']))
+
+tester = RandomTester(T1, 10000, 50)
 
 currentdir = os.path.dirname(os.path.abspath(
                 inspect.getfile(inspect.currentframe())))
@@ -58,7 +89,9 @@ path = os.path.join(currentdir, "dotFiles")
 print("Starting learning...")
 
 # change printPath=None to printPath=path for dot files
-L2 = LearningAlgorithm(T1, O1, printPath=None, maxLoops=4, tablePreciseness=10000, logger=logger, tester=tester)
+L2 = LearningAlgorithm(T1, O1, printPath=None, maxLoops=4,
+    tablePreciseness=10000, logger=logger, tester=tester, outputPurpose=outputExpert,
+    inputPurpose=inputExpert)
 minus, plus = L2.run()
 
 print("Models learned. Check language equivalence...")

examples/tictactoe/completetesting.py

+# Copyright (c) 2015 Michele Volpato
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+# THE SOFTWARE.
+
+from testing.basetesting import AbstractTester
+import logging
+import itertools
+
+# Complete tester.
+class CompleteTicTacToeTester(AbstractTester):
+
+    def __init__(self, teacher, logger=None):
+        # The tester will run test using the teacher
+        self._teacher = teacher
+        # upper bound: 9! (possible plays, also invalid ones)
+        # according to wikipedia
+        # Reduced to games of length 5 (maximum number of X in a bard)
+        self._all_games = set(itertools.permutations(['0','1','2','3','4','5','6','7','8'],5))
+        self._logger = logger or logging.getLogger(__name__)
+
+    # Search a counterexample to teacher ioco model
+    def findCounterexample(self, model):
+        self._teacher.reset()
+        model.reset()
+        ce = ()
+        i = 0
+        for i in self._all_games:
+            ce = ()
+            # Next game
+            self._teacher.reset()
+            model.reset()
+            # play this game, if we receive END, then go to next game.
+            for action in i:
+                ce = ce + (action,)
+                output = self._processInputs((action,))
+                model.move(action)
+                if output not in model.outputs():
+                    self._logger.info("Found a counterexample: "
+                                      + str(ce) + " output: "+str(output))
+
+                    self._teacher.reset()
+                    model.reset()
+                    # return counterexample trace and output obtained by
+                    # testing
+                    return ce, output
+                #elif 'END' in output:
+                #    continue
+                else:
+                    model.move(output)
+                    ce = ce + (output,)
+
+        return None, None
+
+    def _processInputs(self, consecutiveInputs):
+        if consecutiveInputs != ():
+            output = self._teacher.oneOutputForTesting(consecutiveInputs)
+            if output == None:
+                return None
+            return output
+        return self._teacher.output()

examples/tictactoe/learn.py

+# Copyright (c) 2015 Michele Volpato
+#
+# Permission is hereby granted, free of charge, to any person obtaining a copy
+# of this software and associated documentation files (the "Software"), to deal
+# in the Software without restriction, including without limitation the rights
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+# copies of the Software, and to permit persons to whom the Software is
+# furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+# THE SOFTWARE.
+
+import random
+
+random.seed(100)
+
+import os, inspect, sys
+# Include project dir in path
+currentdir = os.path.dirname(os.path.abspath(inspect.getfile(inspect.currentframe())))
+sys.path.append(currentdir)
+
+import socket
+import logging
+
+import itertools
+
+from tictacteacher import TicTacToeTeacher
+from tictacoracle import TicTacToeOracle
+from tictacpurpose import TicTacToeInputPurpose, TicTacToeOutputPurpose
+
+from learning.learning import LearningAlgorithm
+from testing.randomtesting import RandomTester
+from completetesting import CompleteTicTacToeTester
+from systems.implementations import SuspensionAutomaton
+
+import helpers.bisimulation as bi
+
+import csv
+
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+HOST = 'localhost'
+PORT = 29000 # Arbitrary non-privileged port
+
+outputExpert = TicTacToeOutputPurpose()
+inputExpert = TicTacToeInputPurpose()
+
+inputs = set(['0','1','2','3','4','5','6','7','8'])
+# Use a placeholder for outputs
+outputs = outputExpert.allOutputs()
+#outputs = set(itertools.product('XO_', repeat=9))
+
+quiescence = 'delta'
+
+T1 = TicTacToeTeacher(HOST, PORT)
+O1 = TicTacToeOracle(inputs, quiescence)
+
+#tester = RandomTester(T1, 50000, 100)
+tester = CompleteTicTacToeTester(T1)
+
+currentdir = os.path.dirname(os.path.abspath(
+                inspect.getfile(inspect.currentframe())))
+
+path = os.path.join(currentdir, "dotFiles")
+
+print("Starting learning...")
+
+#print(T1.oneOutput(('1')))
+
+L = LearningAlgorithm(T1, O1, printPath=path, maxLoops=4,
+    tablePreciseness=100000, logger=logger, tester=tester, outputPurpose=outputExpert,
+    inputPurpose=inputExpert)
+minus, plus = L.run()
+
+print("Models learned.")
+print("Number of inputs sent to the SUL: " + str(T1.getInputCounter()))
+
+T1.close()
+
+#######################################################################
+# If there is a model learned by LearnLib, load it and run a bisimulation
+# check with minus
+# The file must be converted with learnlib_dot2jtorx_aut.py
+with open("/home/mic/repo/learnLTS/examples/tictactoe/learnLib/TicTacToe.aut", 'r') as csvfile:
+    first = True
+    reader = csv.reader(csvfile, delimiter=';',
+                            quoting=csv.QUOTE_MINIMAL)
+
+    for row in reader:
+        if first:
+            first = False
+            tup = row[0][3:] # remove 'des' from the first line
+            tuple_row = eval(tup)
+            learnLibmodel = SuspensionAutomaton(tuple_row[2],
+                                        inputs.copy(),
+                                        outputs.copy(),
+                                        quiescence,
+                                        False)
+        else:
+            tuple_row = eval(row[0])
+            #try:
+            #    label = eval(tuple_row[1])
+            #except NameError:
+            label = tuple_row[1]
+            learnLibmodel.addTransition(tuple_row[0], label,
+                              tuple_row[2])
+
+print("Models learned. Check language equivalence...")
+
+print("minus bisimilar to LearnLib model: " + str(bi.bisimilar(learnLibmodel,minus,startState1=0, startState2=0, noDelta=True)))
+
+########################################################################

examples/tictactoe/learnLib/README.md

+This is the code used for comparing my tool with learnLib version 0.9.1.
+
+I use this version because of some errors introduced in learnLib by newer versions.
+
+LearnLib code, readme, license and other relevent information can be found at https://github.com/LearnLib/learnlib
+
+How to setup this example:
+In your favorite IDE create a project including the content of the src folder and adding to the path
+the LearnLib library (the content of the lib folder). Everything has been tested in Eclipse Lunawith JavaSE-1.7.
+Adapt the paths to your specific case (around line 178 of TicTacToeLearner.java) and run it to
+learn the SUT. You might need to install Graphviz for the dot program.

examples/tictactoe/learnLib/learnlib_dot2jtorx_aut.py

+#!/usr/bin/env python
+"""
+generates aut for jtorx from dot file learned with learnlib
+  '?COIN_1_1'
+  '!TEA_0_1'
+
+note: dot file uses I for input instead of ? and O for output instead of !
+"""
+# Author: Harco Kuppens
+
+# MOdification by Michele Volpato: removed ? and ! from inputs and outputs
+
+
+import sys, re, pprint  # modules from standard lib (python 2.6 and later)
+
+
+
+def get_lts_from_dotfile(dot_file):
+    """ Get labeled transition system from graphviz dot file
+
+        The dot file:
+           - describes a digraph with labels
+           - encodes the start state with the color='red' attribute
+             note: this corresponds with the highlighted state in learnlib API
+
+        Returns: [start_state,transions]
+        Where :
+           - start_state: start state label
+           - transitions: list of transitions
+
+    """
+
+    start_state='unknown'
+    f=file(dot_file)
+    lines=f.readlines()
+
+
+
+    # find start state
+    #  line in dot: __start0 -> s0;
+    for line in lines:
+        if line.find('->') != -1:
+            if line.find('__start') != -1:
+               start_state=line[line.find('->')+2:].strip(" ;\t\n")
+               break
+
+
+    # get transitions
+    #   line in dot:  s5 -> s5 [label="ARTREG 20013226 / 531"];
+    transitions=[]
+    for line in lines:
+        if line.find('__start') != -1:
+            continue
+        if line.find('->') != -1:
+            transitions.append(line)
+
+    # throw away transitions with the keywords : quiescence or inconsistency or undefined
+    #transitions = [ t for t in transitions if ( 'quiescence' not in t ) and ( 'inconsistency' not in t  )  and ( 'undefined' not in t  )]
+
+    trans_out=[]
+    regexpr_transition=re.compile(r'\s*(\w*)\s*-\>\s*(\w*)\s*\[label=\"(.*)\"\]')
+    regexpr_tag=re.compile(r'<[^>]+>')
+    for transition in transitions:
+        match=regexpr_transition.match(transition)
+        if match:
+            match=match.groups()
+            label=regexpr_tag.sub('',match[2])
+            trans_out.append({
+              'source' : match[0],
+              'target' : match[1],
+              'label': label
+            })
+
+    states=set()
+    for t in trans_out:
+        states.add(t['source'])