Skip to content

Grape

Provides classes and functions for the ATHENA project providing for alternative genome types with corresponding mapping and initialization approaches. Provides alternative selection types for evolutionary algorithm.

Modified code from GRAPE: Grammatical Algorithms in Python for Evolution for use with ATHENA and includes multi-chromosome and LEAP genome implementations

Grammar

BNF Grammar Attributes: non_terminals: list with each non-terminal (NT); start_rule: first non-terminal; nt_rule_size: number of nt rules used for frame size for set of codons in leap mapping and for the number of chromosomes in MCGE production_rules: list with each production rule (PR), which contains in each position: - the PR itself as a string - 'non-terminal' or 'terminal' - the arity (number of NTs in the PR) - production choice label - True, if it is recursive, and False, otherwise - the minimum depth to terminate the mapping of all NTs of this PR n_rules: df

Source code in src/athenage/grape/grape.py 
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
class Grammar:
    """
    BNF Grammar 
    Attributes:
        non_terminals: list with each non-terminal (NT);
        start_rule: first non-terminal;
        nt_rule_size: number of nt rules used for frame size for set of codons in leap mapping 
            and for the number of chromosomes in MCGE
        production_rules: list with each production rule (PR), which contains in each position:
            - the PR itself as a string
            - 'non-terminal' or 'terminal'
            - the arity (number of NTs in the PR)
            - production choice label
            - True, if it is recursive, and False, otherwise
            - the minimum depth to terminate the mapping of all NTs of this PR
        n_rules: df

    """
    def __init__(self, bnf_grammar: str,codon_consumption:str ='eager'):
        """ Initializes Grammar instance by reading BNF grammar string 
        Args:
            bnf_grammar: text describing BNF grammar
            codon_consumption: 'eager' or 'lazy' controlling consumption during mapping

        """
        bnf_grammar = re.sub(r"\s+", " ", bnf_grammar)

        #self.non_terminals = ['<' + term + '>' for term in re.findall(r"\<(\w+)\>\s*::=",bnf_grammar)]
        self.non_terminals = ['<' + term + '>' for term in re.findall(r"\<([\(\)\w,-.]+)\>\s*::=",bnf_grammar)]
        self.start_rule = self.non_terminals[0]
        for i in range(len(self.non_terminals)):
            bnf_grammar = bnf_grammar.replace(self.non_terminals[i] + " ::=", "  ::=")
        rules = bnf_grammar.split("::=")
        del rules[0]
        rules = [item.replace('\n',"") for item in rules]
        rules = [item.replace('\t',"") for item in rules]

        #list of lists (set of production rules for each non-terminal)
        self.production_rules = [i.split('|') for i in rules]
        for i in range(len(self.production_rules)):
            #Getting rid of all leading and trailing whitespaces
            self.production_rules[i] = [item.strip() for item in self.production_rules[i]]
            for j in range(len(self.production_rules[i])):
                #Include in the list the PR itself, NT or T, arity and the production choice label
                #if re.findall(r"\<(\w+)\>",self.production_rules[i][j]):
                if re.findall(r"\<([\(\)\w,-.]+)\>",self.production_rules[i][j]):                    
                    #arity = len(re.findall(r"\<(\w+)\>",self.production_rules[i][j]))
                    arity = len(re.findall(r"\<([\(\)\w,-.]+)\>",self.production_rules[i][j]))
                    self.production_rules[i][j] = [self.production_rules[i][j] , "non-terminal", arity, j]
                else:
                    self.production_rules[i][j] = [self.production_rules[i][j] , "terminal", 0, j] #arity 0
        #number of production rules for each non-terminal
        self.n_rules = [len(list_) for list_ in self.production_rules]

        for i in range(len(self.production_rules)):
            for j in range(len(self.production_rules[i])):
                NTs_to_check_recursiveness = re.findall(r"\<([\(\)\w,-.]+)\>", self.production_rules[i][j][0])
                NTs_to_check_recursiveness = ['<' + item_ + '>' for item_ in NTs_to_check_recursiveness]
                unique_NTs = np.unique(NTs_to_check_recursiveness, return_counts=False) 
                recursive = False
                for NT_to_check in unique_NTs:
                    stack = [self.non_terminals[i]]  
                    if NT_to_check in stack:
                        recursive = True
                        break
                    else:
                        stack.append(NT_to_check)
                        recursive = self.check_recursiveness(NT_to_check, stack)
                        if recursive:
                            break
                        stack.pop()
                self.production_rules[i][j].append(recursive)

        #minimum depth from each non-terminal to terminate the mapping of all symbols
        NT_depth_to_terminate = [None]*len(self.non_terminals)
        #minimum depth from each production rule to terminate the mapping of all symbols
        part_PR_depth_to_terminate = list() #min depth for each non-terminal or terminal to terminate
        isolated_non_terminal = list() #None, if the respective position has a terminal
        #Separating the non-terminals within the same production rule
        for i in range(len(self.production_rules)):
            part_PR_depth_to_terminate.append( list() )
            isolated_non_terminal.append( list() )
            for j in range(len(self.production_rules[i])):
                part_PR_depth_to_terminate[i].append( list() )
                isolated_non_terminal[i].append( list() )
                if self.production_rules[i][j][1] == 'terminal':
                    isolated_non_terminal[i][j].append(None)
                    part_PR_depth_to_terminate[i][j] = 1
                    if not NT_depth_to_terminate[i]:
                        NT_depth_to_terminate[i] = 1
                else:
                    for k in range(self.production_rules[i][j][2]): #arity
                        part_PR_depth_to_terminate[i][j].append( list() )
                        #term = re.findall(r"\<(\w+)\>",self.production_rules[i][j][0])[k]
                        term = re.findall(r"\<([\(\)\w,-.]+)\>",self.production_rules[i][j][0])[k]
                        isolated_non_terminal[i][j].append('<' + term + '>')
        continue_ = True
        while continue_:
            #after filling up NT_depth_to_terminate, we need to run the loop one more time to
            #fill up part_PR_depth_to_terminate, so we check in the beginning
            if None not in NT_depth_to_terminate:
                continue_ = False 
            for i in range(len(self.non_terminals)):
                for j in range(len(self.production_rules)):
                    for k in range(len(self.production_rules[j])):
                        for l in range(len(isolated_non_terminal[j][k])):
                            if self.non_terminals[i] == isolated_non_terminal[j][k][l]:
                                if NT_depth_to_terminate[i]:
                                    if not part_PR_depth_to_terminate[j][k][l]:
                                        part_PR_depth_to_terminate[j][k][l] = NT_depth_to_terminate[i] + 1
                                        if [] not in part_PR_depth_to_terminate[j][k]:
                                            if not NT_depth_to_terminate[j]:
                                                NT_depth_to_terminate[j] = part_PR_depth_to_terminate[j][k][l]
        PR_depth_to_terminate = []
        for i in range(len(part_PR_depth_to_terminate)):
            for j in range(len(part_PR_depth_to_terminate[i])):
                #the min depth to terminate a PR is the max depth within the items of that PR
                if type(part_PR_depth_to_terminate[i][j]) == int:
                    depth_ = part_PR_depth_to_terminate[i][j]
                    PR_depth_to_terminate.append(depth_)
                    self.production_rules[i][j].append(depth_)
                else:
                    depth_ = max(part_PR_depth_to_terminate[i][j])
                    PR_depth_to_terminate.append(depth_)
                    self.production_rules[i][j].append(depth_)

        self.rule_conversion = []
        idx = 0
        self.nt_rule_size = 0
        if codon_consumption == 'lazy':
            for i in range(len(self.non_terminals)):
                if len(self.production_rules[i]) > 1:
                    self.rule_conversion.append(idx)
                    self.nt_rule_size += 1
                    idx += 1
                else:
                    self.rule_conversion.append(-1)
        else: #codon consumption is lazy
            self.rule_conversion=[i for i in range(len(self.non_terminals))]
            self.nt_rule_size = len(self.rule_conversion)


    def check_recursiveness(self, NT: np.str_, stack: list) -> bool:
        """ Determines whether a non-terminal in the grammar is recursive 
        Args:
            NT: non-terminal checked
            stack: tracks non-terminals traversed

        Returns:
            True is recursive, False if not
        """
        idx_NT = self.non_terminals.index(NT)
        for j in range(len(self.production_rules[idx_NT])):
            NTs_to_check_recursiveness = re.findall(r"\<([\(\)\w,-.]+)\>", self.production_rules[idx_NT][j][0])
            NTs_to_check_recursiveness = ['<' + item_ + '>' for item_ in NTs_to_check_recursiveness]
            unique_NTs = np.unique(NTs_to_check_recursiveness, return_counts=False) 
            recursive = False
    #      while unique_NTs.size and not recursive:
            for NT_to_check in unique_NTs:
                if NT_to_check in stack:
                    recursive = True
                    return recursive
                else:
                    stack.append(NT_to_check) #Include the current NT to check it recursively
                    recursive = self.check_recursiveness(NT_to_check, stack)
                    if recursive:
                        return recursive
                    stack.pop() #If the inclusion didn't show recursiveness, remove it before continuing
        return recursive

__init__(bnf_grammar, codon_consumption='eager')

Initializes Grammar instance by reading BNF grammar string Args: bnf_grammar: text describing BNF grammar codon_consumption: 'eager' or 'lazy' controlling consumption during mapping

Source code in src/athenage/grape/grape.py 
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
def __init__(self, bnf_grammar: str,codon_consumption:str ='eager'):
    """ Initializes Grammar instance by reading BNF grammar string 
    Args:
        bnf_grammar: text describing BNF grammar
        codon_consumption: 'eager' or 'lazy' controlling consumption during mapping

    """
    bnf_grammar = re.sub(r"\s+", " ", bnf_grammar)

    #self.non_terminals = ['<' + term + '>' for term in re.findall(r"\<(\w+)\>\s*::=",bnf_grammar)]
    self.non_terminals = ['<' + term + '>' for term in re.findall(r"\<([\(\)\w,-.]+)\>\s*::=",bnf_grammar)]
    self.start_rule = self.non_terminals[0]
    for i in range(len(self.non_terminals)):
        bnf_grammar = bnf_grammar.replace(self.non_terminals[i] + " ::=", "  ::=")
    rules = bnf_grammar.split("::=")
    del rules[0]
    rules = [item.replace('\n',"") for item in rules]
    rules = [item.replace('\t',"") for item in rules]

    #list of lists (set of production rules for each non-terminal)
    self.production_rules = [i.split('|') for i in rules]
    for i in range(len(self.production_rules)):
        #Getting rid of all leading and trailing whitespaces
        self.production_rules[i] = [item.strip() for item in self.production_rules[i]]
        for j in range(len(self.production_rules[i])):
            #Include in the list the PR itself, NT or T, arity and the production choice label
            #if re.findall(r"\<(\w+)\>",self.production_rules[i][j]):
            if re.findall(r"\<([\(\)\w,-.]+)\>",self.production_rules[i][j]):                    
                #arity = len(re.findall(r"\<(\w+)\>",self.production_rules[i][j]))
                arity = len(re.findall(r"\<([\(\)\w,-.]+)\>",self.production_rules[i][j]))
                self.production_rules[i][j] = [self.production_rules[i][j] , "non-terminal", arity, j]
            else:
                self.production_rules[i][j] = [self.production_rules[i][j] , "terminal", 0, j] #arity 0
    #number of production rules for each non-terminal
    self.n_rules = [len(list_) for list_ in self.production_rules]

    for i in range(len(self.production_rules)):
        for j in range(len(self.production_rules[i])):
            NTs_to_check_recursiveness = re.findall(r"\<([\(\)\w,-.]+)\>", self.production_rules[i][j][0])
            NTs_to_check_recursiveness = ['<' + item_ + '>' for item_ in NTs_to_check_recursiveness]
            unique_NTs = np.unique(NTs_to_check_recursiveness, return_counts=False) 
            recursive = False
            for NT_to_check in unique_NTs:
                stack = [self.non_terminals[i]]  
                if NT_to_check in stack:
                    recursive = True
                    break
                else:
                    stack.append(NT_to_check)
                    recursive = self.check_recursiveness(NT_to_check, stack)
                    if recursive:
                        break
                    stack.pop()
            self.production_rules[i][j].append(recursive)

    #minimum depth from each non-terminal to terminate the mapping of all symbols
    NT_depth_to_terminate = [None]*len(self.non_terminals)
    #minimum depth from each production rule to terminate the mapping of all symbols
    part_PR_depth_to_terminate = list() #min depth for each non-terminal or terminal to terminate
    isolated_non_terminal = list() #None, if the respective position has a terminal
    #Separating the non-terminals within the same production rule
    for i in range(len(self.production_rules)):
        part_PR_depth_to_terminate.append( list() )
        isolated_non_terminal.append( list() )
        for j in range(len(self.production_rules[i])):
            part_PR_depth_to_terminate[i].append( list() )
            isolated_non_terminal[i].append( list() )
            if self.production_rules[i][j][1] == 'terminal':
                isolated_non_terminal[i][j].append(None)
                part_PR_depth_to_terminate[i][j] = 1
                if not NT_depth_to_terminate[i]:
                    NT_depth_to_terminate[i] = 1
            else:
                for k in range(self.production_rules[i][j][2]): #arity
                    part_PR_depth_to_terminate[i][j].append( list() )
                    #term = re.findall(r"\<(\w+)\>",self.production_rules[i][j][0])[k]
                    term = re.findall(r"\<([\(\)\w,-.]+)\>",self.production_rules[i][j][0])[k]
                    isolated_non_terminal[i][j].append('<' + term + '>')
    continue_ = True
    while continue_:
        #after filling up NT_depth_to_terminate, we need to run the loop one more time to
        #fill up part_PR_depth_to_terminate, so we check in the beginning
        if None not in NT_depth_to_terminate:
            continue_ = False 
        for i in range(len(self.non_terminals)):
            for j in range(len(self.production_rules)):
                for k in range(len(self.production_rules[j])):
                    for l in range(len(isolated_non_terminal[j][k])):
                        if self.non_terminals[i] == isolated_non_terminal[j][k][l]:
                            if NT_depth_to_terminate[i]:
                                if not part_PR_depth_to_terminate[j][k][l]:
                                    part_PR_depth_to_terminate[j][k][l] = NT_depth_to_terminate[i] + 1
                                    if [] not in part_PR_depth_to_terminate[j][k]:
                                        if not NT_depth_to_terminate[j]:
                                            NT_depth_to_terminate[j] = part_PR_depth_to_terminate[j][k][l]
    PR_depth_to_terminate = []
    for i in range(len(part_PR_depth_to_terminate)):
        for j in range(len(part_PR_depth_to_terminate[i])):
            #the min depth to terminate a PR is the max depth within the items of that PR
            if type(part_PR_depth_to_terminate[i][j]) == int:
                depth_ = part_PR_depth_to_terminate[i][j]
                PR_depth_to_terminate.append(depth_)
                self.production_rules[i][j].append(depth_)
            else:
                depth_ = max(part_PR_depth_to_terminate[i][j])
                PR_depth_to_terminate.append(depth_)
                self.production_rules[i][j].append(depth_)

    self.rule_conversion = []
    idx = 0
    self.nt_rule_size = 0
    if codon_consumption == 'lazy':
        for i in range(len(self.non_terminals)):
            if len(self.production_rules[i]) > 1:
                self.rule_conversion.append(idx)
                self.nt_rule_size += 1
                idx += 1
            else:
                self.rule_conversion.append(-1)
    else: #codon consumption is lazy
        self.rule_conversion=[i for i in range(len(self.non_terminals))]
        self.nt_rule_size = len(self.rule_conversion)

check_recursiveness(NT, stack)

Determines whether a non-terminal in the grammar is recursive Args: NT: non-terminal checked stack: tracks non-terminals traversed

Returns:

Type Description
bool

True is recursive, False if not
Source code in src/athenage/grape/grape.py 
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
def check_recursiveness(self, NT: np.str_, stack: list) -> bool:
    """ Determines whether a non-terminal in the grammar is recursive 
    Args:
        NT: non-terminal checked
        stack: tracks non-terminals traversed

    Returns:
        True is recursive, False if not
    """
    idx_NT = self.non_terminals.index(NT)
    for j in range(len(self.production_rules[idx_NT])):
        NTs_to_check_recursiveness = re.findall(r"\<([\(\)\w,-.]+)\>", self.production_rules[idx_NT][j][0])
        NTs_to_check_recursiveness = ['<' + item_ + '>' for item_ in NTs_to_check_recursiveness]
        unique_NTs = np.unique(NTs_to_check_recursiveness, return_counts=False) 
        recursive = False
#      while unique_NTs.size and not recursive:
        for NT_to_check in unique_NTs:
            if NT_to_check in stack:
                recursive = True
                return recursive
            else:
                stack.append(NT_to_check) #Include the current NT to check it recursively
                recursive = self.check_recursiveness(NT_to_check, stack)
                if recursive:
                    return recursive
                stack.pop() #If the inclusion didn't show recursiveness, remove it before continuing
    return recursive

Individual

A GE individual.

Source code in src/athenage/grape/grape.py 
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
class Individual:
    """
    A GE individual.
    """

    def __init__(self, genome, grammar, max_depth, codon_consumption):
        """
        """

        self.genome = genome
        if codon_consumption == 'lazy':
            self.phenotype, self.nodes, self.depth, \
            self.used_codons, self.invalid, self.n_wraps, \
            self.structure = mapper_lazy(genome, grammar, max_depth)
        elif codon_consumption == 'eager':
            self.phenotype, self.nodes, self.depth, \
            self.used_codons, self.invalid, self.n_wraps, \
            self.structure = mapper_eager(genome, grammar, max_depth)
#             self.structure = mapper_eager_opt(genome, grammar, max_depth)


        else:
            raise ValueError("Unknown mapper")

__init__(genome, grammar, max_depth, codon_consumption)

Source code in src/athenage/grape/grape.py 
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
    def __init__(self, genome, grammar, max_depth, codon_consumption):
        """
        """

        self.genome = genome
        if codon_consumption == 'lazy':
            self.phenotype, self.nodes, self.depth, \
            self.used_codons, self.invalid, self.n_wraps, \
            self.structure = mapper_lazy(genome, grammar, max_depth)
        elif codon_consumption == 'eager':
            self.phenotype, self.nodes, self.depth, \
            self.used_codons, self.invalid, self.n_wraps, \
            self.structure = mapper_eager(genome, grammar, max_depth)
#             self.structure = mapper_eager_opt(genome, grammar, max_depth)


        else:
            raise ValueError("Unknown mapper")

create_genome(genome_type, n_rules, codon_size)

Create and return appropriate Genome

Parameters:

Name Type Description Default
genome_type str

Type to create (leap,mcge,standard)

 required
n_rules int

number of rules in the grammar used to map this genome

 required
codon_size int

maximum value for a codon

required

Returns:

Type Description
Genome

new Genome
Source code in src/athenage/grape/grape.py 
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
def create_genome(genome_type:str, n_rules:int, codon_size:int) -> "grape.grape.Genome":
    """Create and return appropriate Genome 

    Args: 
        genome_type: Type to create (leap,mcge,standard)
        n_rules: number of rules in the grammar used to map this genome
        codon_size: maximum value for a codon 

    Returns:
        new Genome
    """
    if genome_type == 'leap':
        return LeapGenome(n_rules, codon_size)
    elif genome_type == 'mcge':
        return MCGEGenome(n_rules, codon_size)
    else:  
        return Genome(n_rules, codon_size)

crossover_block(parent1, parent2, bnf_grammar, max_depth, codon_consumption, genome_representation='list', max_genome_length=None)

Two point crossover for GE individuals restricted to locations where identical rule selections are made Starts at matching point and ends with the expression tree portion terminated from that point.

Parameters:

Name Type Description Default
parent1 Individual

individual to cross

 required
parent2 Individual

second individual to cross

 required
bnf_grammar Grammar

BNF grammar

required
max_depth int

maximum depth allowed for a tree during mapping

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required
genome_representation str

either list or numpy

 'list'
max_genome_length int

maximum allowed length of genome when not None

 None

Returns:

Name Type Description
new_ind0 Individual

new individual resulting from crossover
new_ind1 Individual

second new individual resulting from crossover
Source code in src/athenage/grape/grape.py 
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
def crossover_block(parent1: 'deap.creator.Individual', parent2: 'deap.creator.Individual', 
                       bnf_grammar: 'grape.grape.Grammar', max_depth: int, codon_consumption:str, 
                    genome_representation:str='list', 
                    max_genome_length:int=None) -> tuple['deap.creator.Individual','deap.creator.Individual']:
    """Two point crossover for GE individuals restricted to locations where identical rule selections are made
    Starts at matching point and ends with the expression tree portion terminated from that point.

    Args:
        parent1: individual to cross
        parent2: second individual to cross
        bnf_grammar: BNF grammar 
        max_depth: maximum depth allowed for a tree during mapping
        codon_consumption: type of consumption when mapping (ie. lazy or eager)
        genome_representation: either list or numpy
        max_genome_length: maximum allowed length of genome when not None

    Returns:
        new_ind0: new individual resulting from crossover
        new_ind1: second new individual resulting from crossover
    """ 

    # restrict crossover to effective genome when individual is valid
    if parent1.invalid: #used_codons = 0
        pos1 = parent1.genome.all_cross_loc()
    else:
        pos1 = parent1.genome.effective_cross_loc()
    rule1_idx = parent1.genome.rule_used(pos1[0])

    pos1_end = submap(parent1, parent1.genome, bnf_grammar, pos1, rule1_idx, codon_consumption)

    if parent2.invalid:
        pos2 = parent2.genome.all_cross_loc()
        check_len = parent2.genome.total_codons()
    else:
        pos2 = parent2.genome.effective_cross_loc()
        check_len = parent2.genome.used_codons()

    # start with selected position and then move up or down list to find match
    if random.random() < 0.5:
        range1 = range(pos2[0],check_len, 1)
        range2 = range(pos2[0]-1, -1, -1)
    else:
        range1 = range(pos2[0], -1, -1)
        range2 = range(pos2[0]+1,check_len,1)

    found = False
    for i in range1:
        if parent2.genome.rule_used(i) == rule1_idx:
            found = True
            pos2 = [i]
            break
    if not found:
        for i in range2:
            if parent2.genome.rule_used(i) == rule1_idx:
                found = True
                pos2 = [i]
                break

    rule2_idx = parent2.genome.rule_used(pos2[0])

    pos2_end = submap(parent2, parent2.genome, bnf_grammar, pos2, rule2_idx, codon_consumption)

    new_genome1, new_genome2 = parent1.genome.crossover_twopoint(parent2.genome, 
                                                                 [pos1[0],pos1_end], [pos2[0],pos2_end])

    new_ind0 = reMap(parent1, new_genome1, bnf_grammar, max_depth, codon_consumption)
    new_ind1 = reMap(parent2, new_genome2, bnf_grammar, max_depth, codon_consumption)

    if new_ind0.depth > max_depth:
        new_ind0.invalid = True
    if new_ind1.depth > max_depth:
        new_ind1.invalid = True

    if max_genome_length:
        if len(new_ind0.genome) > max_genome_length:
            new_ind0.invalid = True
        if len(new_ind1.genome) > max_genome_length:
            new_ind1.invalid = True

    del new_ind0.fitness.values, new_ind1.fitness.values
    return new_ind0, new_ind1

crossover_match(parent1, parent2, bnf_grammar, max_depth, codon_consumption, genome_representation='list', max_genome_length=None)

One point crossover for GE individuals restricted to locations where identical rule selections are made

Parameters:

Name Type Description Default
parent1 Individual

individual to cross

 required
parent2 Individual

second individual to cross

 required
bnf_grammar Grammar

BNF grammar

required
max_depth int

maximum depth allowed for a tree during mapping

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required
genome_representation str

either list or numpy

 'list'
max_genome_length int

maximum allowed length of genome when not None

 None

Returns:

Name Type Description
new_ind0 Individual

new individual resulting from crossover
new_ind1 Individual

second new individual resulting from crossover
Source code in src/athenage/grape/grape.py 
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
def crossover_match(parent1: 'deap.creator.Individual', parent2: 'deap.creator.Individual', 
                       bnf_grammar: 'grape.grape.Grammar', max_depth: int, codon_consumption:str, 
                    genome_representation:str='list', 
                    max_genome_length:int=None) -> tuple['deap.creator.Individual','deap.creator.Individual']:
    """One point crossover for GE individuals restricted to locations where identical rule selections are made

    Args:
        parent1: individual to cross
        parent2: second individual to cross
        bnf_grammar: BNF grammar 
        max_depth: maximum depth allowed for a tree during mapping
        codon_consumption: type of consumption when mapping (ie. lazy or eager)
        genome_representation: either list or numpy
        max_genome_length: maximum allowed length of genome when not None

    Returns:
        new_ind0: new individual resulting from crossover
        new_ind1: second new individual resulting from crossover
    """ 

    # restrict crossover to effective genome when individual is valid
    if parent1.invalid: #used_codons = 0
        pos1 = parent1.genome.all_cross_loc()
    else:
        pos1 = parent1.genome.effective_cross_loc()
    rule1_idx = parent1.genome.rule_used(pos1[0])

    if parent2.invalid:
        pos2 = parent2.genome.all_cross_loc()
        check_len = parent2.genome.used_codons()
    else:
        pos2 = parent2.genome.effective_cross_loc()
        check_len = parent2.genome.total_codons()

    # start with selected position and then move up or down list to find match
    if random.random() < 0.5:
        range1 = range(pos2[0],check_len, 1)
        range2 = range(pos2[0]-1, -1, -1)
    else:
        range1 = range(pos2[0], -1, -1)
        range2 = range(pos2[0]+1,check_len,1)

    found = False
    for i in range1:
        if parent2.genome.rule_used(i) == rule1_idx:
            found = True
            pos2 = [i]
            break
    if not found:
        for i in range2:
            if parent2.genome.rule_used(i) == rule1_idx:
                found = True
                pos2 = [i]
                break

    new_genome1, new_genome2 = parent1.genome.crossover_onepoint(parent2.genome, pos1, pos2)

    new_ind0 = reMap(parent1, new_genome1, bnf_grammar, max_depth, codon_consumption)
    new_ind1 = reMap(parent2, new_genome2, bnf_grammar, max_depth, codon_consumption)

    if new_ind0.depth > max_depth:
        new_ind0.invalid = True
    if new_ind1.depth > max_depth:
        new_ind1.invalid = True

    if max_genome_length:
        if len(new_ind0.genome) > max_genome_length:
            new_ind0.invalid = True
        if len(new_ind1.genome) > max_genome_length:
            new_ind1.invalid = True

    del new_ind0.fitness.values, new_ind1.fitness.values
    return new_ind0, new_ind1

crossover_onepoint(parent1, parent2, bnf_grammar, max_depth, codon_consumption, genome_representation='list', max_genome_length=None)

One point crossover for GE individuals

Parameters:

Name Type Description Default
parent1 Individual

individual to cross

 required
parent2 Individual

second individual to cross

 required
bnf_grammar Grammar

BNF grammar

required
max_depth int

maximum depth allowed for a tree during mapping

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required
genome_representation str

either list or numpy

 'list'
max_genome_length int

maximum allowed length of genome when not None

 None

Returns:

Name Type Description
new_ind0 Individual

new individual resulting from crossover
new_ind1 Individual

second new individual resulting from crossover
Source code in src/athenage/grape/grape.py 
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
def crossover_onepoint(parent1: 'deap.creator.Individual', parent2: 'deap.creator.Individual', 
                       bnf_grammar: 'grape.grape.Grammar', max_depth: int, codon_consumption:str, 
                    genome_representation:str='list', 
                    max_genome_length:int=None) -> tuple['deap.creator.Individual','deap.creator.Individual']:
    """One point crossover for GE individuals

    Args:
        parent1: individual to cross
        parent2: second individual to cross
        bnf_grammar: BNF grammar 
        max_depth: maximum depth allowed for a tree during mapping
        codon_consumption: type of consumption when mapping (ie. lazy or eager)
        genome_representation: either list or numpy
        max_genome_length: maximum allowed length of genome when not None

    Returns:
        new_ind0: new individual resulting from crossover
        new_ind1: second new individual resulting from crossover
    """ 
    # restrict crossover to effective genome when individual is valid
    if parent1.invalid: #used_codons = 0
        # possible_crossover_codons0 = len(parent0.genome)
        pos1 = parent1.genome.all_cross_loc()
    else:
        pos1 = parent1.genome.effective_cross_loc()
    if parent2.invalid:
        pos2 = parent2.genome.all_cross_loc()
    else:
        pos2 = parent2.genome.effective_cross_loc()


    new_genome1, new_genome2 = parent1.genome.crossover_onepoint(parent2.genome, pos1, pos2)

    new_ind0 = reMap(parent1, new_genome1, bnf_grammar, max_depth, codon_consumption)
    new_ind1 = reMap(parent2, new_genome2, bnf_grammar, max_depth, codon_consumption)

    if new_ind0.depth > max_depth:
        new_ind0.invalid = True
    if new_ind1.depth > max_depth:
        new_ind1.invalid = True

    if max_genome_length:
        if len(new_ind0.genome) > max_genome_length:
            new_ind0.invalid = True
        if len(new_ind1.genome) > max_genome_length:
            new_ind1.invalid = True

    del new_ind0.fitness.values, new_ind1.fitness.values
    return new_ind0, new_ind1

mapper_eager(genome, grammar, max_depth)

Maps GE genome. Uses eager mapping so that every nonterminal in the grammar consumes a codon in the genome

Parameters:

Name Type Description Default
genome Genome

list of codons in genome

 required
grammar grammar

BNF grammar used to create individual phenotype

 required
max_depth int

maximum depth allowed for a tree during mapping

 required

Returns:

Name Type Description
phenotype str

mapped phenotype

nodes int

number of nodes in tree
depth int

max depth of tree used in mapping
used_codons int

number of codons consumed in mapping
invalid bool

True if valid phenotype not constructed
list

int for number of wraps used (0 for this implementation as wrapping not used)
structure tuple[str, int, int, int, bool, list]

list of production rules chosen in mapping, used to monitor population diversity
Source code in src/athenage/grape/grape.py 
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
def mapper_eager(genome: "grape.grape.Genome", grammar: "grape.grape.grammar",
                  max_depth: int) -> tuple[str, int, int, int, bool, list]:
    """Maps GE  genome. Uses eager mapping so that 
    every nonterminal in the grammar consumes a codon in the genome

    Args:
        genome: list of codons in genome
        grammar: BNF grammar used to create individual phenotype
        max_depth: maximum depth allowed for a tree during mapping


    Returns:
        phenotype: mapped phenotype 
        nodes: number of nodes in tree
        depth: max depth of tree used in mapping
        used_codons: number of codons consumed in mapping
        invalid: True if valid phenotype not constructed
        int for number of wraps used (0 for this implementation as wrapping not used)
        structure: list of production rules chosen in mapping, 
            used to monitor population diversity
    """

    # idx_genome = 0
    genome.reset_map_index()
    phenotype = grammar.start_rule
    next_NT = re.search(r"\<(\w+)\>",phenotype).group()
    n_starting_NTs = len([term for term in re.findall(r"\<(\w+)\>",phenotype)])
    list_depth = [1]*n_starting_NTs #it keeps the depth of each branch
    idx_depth = 0
    nodes = 0
    structure = []

    # while next_NT and idx_genome < len(genome):
    while next_NT:
        NT_index = grammar.non_terminals.index(next_NT)
        if genome.consumed(NT_index):
            break
        # index_production_chosen = genome[idx_genome] % grammar.n_rules[NT_index]
        index_production_chosen = genome.get_next_codon(NT_index) % grammar.n_rules[NT_index]
        structure.append(index_production_chosen)
        phenotype = phenotype.replace(next_NT, grammar.production_rules[NT_index][index_production_chosen][0], 1)
        list_depth[idx_depth] += 1
        if list_depth[idx_depth] > max_depth:
            break
        if grammar.production_rules[NT_index][index_production_chosen][2] == 0: #arity 0 (T)
            idx_depth += 1
            nodes += 1
        elif grammar.production_rules[NT_index][index_production_chosen][2] == 1: #arity 1 (PR with one NT)
            pass        
        else: #it is a PR with more than one NT
            arity = grammar.production_rules[NT_index][index_production_chosen][2]
            if idx_depth == 0:
                list_depth = [list_depth[idx_depth],]*arity + list_depth[idx_depth+1:]
            else:
                list_depth = list_depth[0:idx_depth] + [list_depth[idx_depth],]*arity + list_depth[idx_depth+1:]

        next_ = re.search(r"\<(\w+)\>",phenotype)
        if next_:
            next_NT = next_.group()
        else:
            next_NT = None
        # idx_genome += 1

    if next_NT:
        invalid = True
        used_codons = 0
    else:
        invalid = False
        used_codons = genome.used_codons()

    depth = max(list_depth)

    return phenotype, nodes, depth, used_codons, invalid, 0, structure

mapper_lazy(genome, grammar, max_depth)

Uses lazy mapping so that only nonterminals with more than one option in the grammar consume a codon in the genome

Parameters:

Name Type Description Default
genome list

list of codons in genome

 required
grammar grammar

BNF grammar used to create individual phenotype

 required
max_depth int

maximum depth allowed for a tree during mapping

 required

Returns:

Name Type Description
phenotype str
nodes int

number of nodes in tree
depth int

max depth of tree used in mapping
used_codons int

number of codons consumed in mapping
invalid bool

True if valid phenotype not constructed
list

int for number of wraps used (0 for this implementation as wrapping not used)
structure tuple[str, int, int, int, bool, list]

list of production rules chosen in mapping, used to monitor population diversity
Source code in src/athenage/grape/grape.py 
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
def mapper_lazy(genome: list, grammar: "grape.grape.grammar",
                  max_depth: int) -> tuple[str, int, int, int, bool, list]:
    """ Uses lazy mapping so that 
    only nonterminals with more than one option in the grammar 
    consume a codon in the genome

    Args:
        genome: list of codons in genome
        grammar: BNF grammar used to create individual phenotype
        max_depth: maximum depth allowed for a tree during mapping


    Returns:
        phenotype: 
        nodes: number of nodes in tree
        depth: max depth of tree used in mapping
        used_codons: number of codons consumed in mapping
        invalid: True if valid phenotype not constructed
        int for number of wraps used (0 for this implementation as wrapping not used)
        structure: list of production rules chosen in mapping, 
            used to monitor population diversity
    """  

    genome.reset_map_index()
    phenotype = grammar.start_rule
    next_NT = re.search(r"\<(\w+)\>",phenotype).group()
    n_starting_NTs = len([term for term in re.findall(r"\<(\w+)\>",phenotype)])
    list_depth = [1]*n_starting_NTs #it keeps the depth of each branch
    idx_depth = 0
    nodes = 0
    structure = []

    while next_NT:
        NT_index = grammar.non_terminals.index(next_NT)
        if genome.consumed(NT_index):
            break        
        if grammar.n_rules[NT_index] == 1: #there is a single PR for this non-terminal
            index_production_chosen = 0        
        else: #we consume one codon, and add the index to the structure
            index_production_chosen = genome.get_next_codon(NT_index) % grammar.n_rules[NT_index]
            structure.append(index_production_chosen)
            # idx_genome += 1

        phenotype = phenotype.replace(next_NT, grammar.production_rules[NT_index][index_production_chosen][0], 1)
        list_depth[idx_depth] += 1
        if list_depth[idx_depth] > max_depth:
            break
        if grammar.production_rules[NT_index][index_production_chosen][2] == 0: #arity 0 (T)
            idx_depth += 1
            nodes += 1
        elif grammar.production_rules[NT_index][index_production_chosen][2] == 1: #arity 1 (PR with one NT)
            pass        
        else: #it is a PR with more than one NT
            arity = grammar.production_rules[NT_index][index_production_chosen][2]
            if idx_depth == 0:
                list_depth = [list_depth[idx_depth],]*arity + list_depth[idx_depth+1:]
            else:
                list_depth = list_depth[0:idx_depth] + [list_depth[idx_depth],]*arity + list_depth[idx_depth+1:]

        next_ = re.search(r"\<(\w+)\>",phenotype)
        if next_:
            next_NT = next_.group()
        else:
            next_NT = None


    if next_NT:
        invalid = True
        used_codons = 0
    else:
        invalid = False
        used_codons = genome.used_codons()

    depth = max(list_depth)

    return phenotype, nodes, depth, used_codons, invalid, 0, structure

mutation_int_flip_per_codon(ind, mut_probability, codon_size, bnf_grammar, max_depth, codon_consumption, max_genome_length=None)

Mutation operator for GE genome. Each codon within the effective used codons of the genome is checked

Parameters:

Name Type Description Default
ind Individual

individual to mutate

 required
mut_probability float

chance for each codon to mutate

 required
codon_size int

maximum value for a codon

 required
bnf_grammar Grammar

BNF grammar

required
max_depth int

maximum depth allowed for a tree during mapping

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required
max_genome_length int

maximum allowed length of genome when not None

 None

Returns:

Name Type Description
new_ind Individual

new individual incorporating any mutations
Source code in src/athenage/grape/grape.py 
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
def mutation_int_flip_per_codon(ind:'deap.creator.Individual', mut_probability:float, 
                                codon_size:int , bnf_grammar:'grape.grape.Grammar', max_depth:int, 
                                codon_consumption:str, 
                                max_genome_length:int=None) -> 'deap.creator.Individual':
    """Mutation operator for GE genome. Each codon within the effective used codons
    of the genome is checked

    Args:
        ind: individual to mutate
        mut_probability: chance for each codon to mutate
        codon_size: maximum value for a codon
        bnf_grammar: BNF grammar 
        max_depth: maximum depth allowed for a tree during mapping
        codon_consumption: type of consumption when mapping (ie. lazy or eager)
        max_genome_length: maximum allowed length of genome when not None

    Returns:
        new_ind: new individual incorporating any mutations
    """ 

    # Operation mutation within the effective part of the genome
    if ind.invalid: #used_codons = 0
        possible_mutation_codons = ind.genome.total_codons()
    else:
        possible_mutation_codons = min(ind.genome.used_codons(), ind.genome.total_codons())

    mutated_ = False

    for i in range(possible_mutation_codons):
        if random.random() < mut_probability:
            ind.genome.set_codon(i, random.randint(0,codon_size), ind.invalid)
            mutated_ = True

    if mutated_:
        new_ind = reMap(ind, ind.genome, bnf_grammar, max_depth, codon_consumption)
    else:
        new_ind = ind

    if new_ind.depth > max_depth:
        new_ind.invalid = True

    if max_genome_length:
        if len(new_ind.genome) > max_genome_length:
            new_ind.invalid = True

    if mutated_:
        del new_ind.fitness.values
    return new_ind,

random_initialization(ind_class, pop_size, bnf_grammar, min_init_genome_length, max_init_genome_length, max_init_depth, codon_size, codon_consumption, genome_representation, genome_type)

Randomly generated genomes that may be invalid when mapped

Parameters:

Name Type Description Default
ind_class MetaCreator

Deap creator

 required
pop_size int

number of individuals to create

 required
bnf_grammar Grammar

BNF grammar

required
min_init_genome_length int

minimum number of codons in genome

 required
max_init_genome_length int

maximum number of codons in genome

 required
max_init_depth int

maximum depth for tree in mapping genome

 required
codon_size int

maximum to be stored in codon

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required
genome_representation str

either list or numpy

 required
genome_type str

standard/mcge/leap

 required

Returns:

Name Type Description
population list

list of individuals
Source code in src/athenage/grape/grape.py 
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
def random_initialization(ind_class: 'deap.creator.MetaCreator', pop_size:int, 
                          bnf_grammar:'grape.grape.Grammar', 
                          min_init_genome_length:int, max_init_genome_length:int,
                          max_init_depth:int, codon_size:int, codon_consumption:str,
                          genome_representation:str, genome_type:str) -> list:
    """Randomly generated genomes that may be invalid when mapped

    Args:
        ind_class: Deap creator
        pop_size: number of individuals to create
        bnf_grammar: BNF grammar 
        min_init_genome_length: minimum number of codons in genome
        max_init_genome_length: maximum number of codons in genome
        max_init_depth: maximum depth for tree in mapping genome
        codon_size: maximum to be stored in codon
        codon_consumption: type of consumption when mapping (ie. lazy or eager)
        genome_representation: either list or numpy
        genome_type: standard/mcge/leap


    Returns:
        population: list of individuals
    """ 

    population = []

    for i in range(pop_size):
        genome = create_genome(genome_type,bnf_grammar.nt_rule_size, codon_size)
        init_genome_length = random.randint(min_init_genome_length, max_init_genome_length)
        genome.random_fill(init_genome_length)
        genome.finalize()
        ind = ind_class(genome, bnf_grammar, max_init_depth, codon_consumption)
        population.append(ind)

    if genome_representation == 'list':
        return population
    else:
        raise ValueError("Unkonwn genome representation")

reMap(ind, genome, bnf_grammar, max_tree_depth, codon_consumption)

Maps GE genome in individual to produce new phenotype

Parameters:

Name Type Description Default
ind Individual

individual to mutate

 required
bnf_grammar Grammar

BNF grammar

required
max_tree_depth int

maximum depth allowed for a tree during mapping

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required

Returns:

Name Type Description
new_ind Individual

new individual with updating phenotype and other mapping information
Source code in src/athenage/grape/grape.py 
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
def reMap(ind: 'deap.creator.Individual', genome: list, 
          bnf_grammar:'grape.grape.Grammar', max_tree_depth:int, 
          codon_consumption:str) -> 'deap.creator.Individual':
    """Maps GE genome in individual to produce new phenotype

    Args:
        ind: individual to mutate
        bnf_grammar: BNF grammar 
        max_tree_depth: maximum depth allowed for a tree during mapping
        codon_consumption: type of consumption when mapping (ie. lazy or eager)

    Returns:
        new_ind: new individual with updating phenotype and other mapping information
    """ 

    ind.genome = genome
    if codon_consumption == 'lazy':
        ind.phenotype, ind.nodes, ind.depth, \
        ind.used_codons, ind.invalid, ind.n_wraps, \
        ind.structure = mapper_lazy(genome, bnf_grammar, max_tree_depth)
    elif codon_consumption == 'eager':
        ind.phenotype, ind.nodes, ind.depth, \
        ind.used_codons, ind.invalid, ind.n_wraps, \
        ind.structure = mapper_eager(genome, bnf_grammar, max_tree_depth)
    else:
        raise ValueError("Unknown mapper")

    return ind

replace_nth(string, substring, new_substring, nth)

utility function to replace nth instance of a substring

Parameters:

Name Type Description Default
string str

original str

 required
substring str

substring to replace

 required
new_substring str

new substring to use

 required
nth int

instance of occurrence of substring to replace

 required

Returns; updated string

Source code in src/athenage/grape/grape.py 
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
def replace_nth(string:str, substring:str, new_substring:str, nth:int) -> str:
    """ utility function to replace nth instance of a substring

    Args:
        string: original str
        substring: substring to replace
        new_substring: new substring to use
        nth: instance of occurrence of substring to replace

    Returns;
        updated string

    """
    find = string.find(substring)
    i = find != -1
    while find != -1 and i != nth:
        find = string.find(substring, find + 1)
        i += 1
    if i == nth:
        return string[:find] + new_substring + string[find+len(substring):]
    return string

selAthenaRoulette(individuals, k, fit_attr='fitness')

Matches the original C++ ATHENA roulette selection method

Parameters:

Name Type Description Default
individuals list

A list of individuals to select from.

 required
k int

The number of individuals to select.

 required
fit_attr str

The attribute of individuals to use as selection criterion

 'fitness'

Returns:

Type Description
list

A list containing the k selected individuals
Source code in src/athenage/grape/grape.py 
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
def selAthenaRoulette(individuals: list, k:int, fit_attr:str="fitness") -> list:
    """Matches the original C++ ATHENA roulette selection method

    Args:
        individuals: A list of individuals to select from.
        k: The number of individuals to select.
        fit_attr: The attribute of individuals to use as selection criterion

    Returns: 
        A list containing the k selected individuals
    """
    # sort the population in ascending order
    individuals = sorted(individuals, key=attrgetter(fit_attr), reverse=True)

    psum=[]
    n = len(individuals)

    # set the probabilities for each i ndividual to be selected
    # if all identical then make all probabilities equal
    if individuals[0].fitness.values[0] == individuals[-1].fitness.values[0]:
        psum = [i.fitness/float(n) for i in individuals]
    else:
        maxfit = individuals[-1].fitness.values[0]
        minfit = individuals[0].fitness.values[0]
        # set first probability
        psum.append(-individuals[0].fitness.values[0] + maxfit + minfit)

        for i,ind in enumerate(individuals[1:], start=1):
            psum.append(-ind.fitness.values[0] + minfit + maxfit + psum[i-1])

        for i in range(n):
            psum[i] /= psum[-1]

    # select individuals
    chosen=[]
    for i in range(k):
        #use a binary search 
        cutoff = random.random()
        lowerbound = 0
        upperbound = n-1

        while(upperbound >= lowerbound):
            indindex=lowerbound + (upperbound-lowerbound)//2
            if psum[indindex] > cutoff:
                upperbound = indindex-1
            else:
                lowerbound = indindex+1

        lowerbound = min(n-1, lowerbound) 
        lowerbound = max(0, lowerbound)

        chosen.append(individuals[lowerbound])    

    return chosen

selAutoEpsilonLexicase(individuals, k)

Adapted from DEAP selAutomaticEpsilonLexicase Returns an individual that does the best on the fitness cases when considered one at a time in random order. https://push-language.hampshire.edu/uploads/default/original/1X/35c30e47ef6323a0a949402914453f277fb1b5b0.pdf Implemented lambda_epsilon_y implementation.

Args; individuals: A list of individuals to select from. k: The number of individuals to select.

Returns:

Type Description
list

A list of selected individuals.
Source code in src/athenage/grape/grape.py 
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
def selAutoEpsilonLexicase(individuals: list, k: int) -> list:
    """
    Adapted from DEAP selAutomaticEpsilonLexicase
    Returns an individual that does the best on the fitness cases when considered one at a
    time in random order.
    https://push-language.hampshire.edu/uploads/default/original/1X/35c30e47ef6323a0a949402914453f277fb1b5b0.pdf
    Implemented lambda_epsilon_y implementation.

    Args;
        individuals: A list of individuals to select from.
        k: The number of individuals to select.

    Returns: 
        A list of selected individuals.
    """

    selected_individuals = []
    nscores =  len(individuals[0].ptscores)

    for i in range(k):
        candidates = individuals
        cases = list(range(nscores))
        random.shuffle(cases)

        while len(cases) > 0 and len(candidates) > 1:
            errors_for_this_case = np.array([x.ptscores[cases[0]] for x in candidates])

            # check if all candidates are nan for this case
            if np.count_nonzero(np.isnan(errors_for_this_case)) == len(candidates):
                break
            nan_mask = np.isnan(errors_for_this_case)

            median_val = np.median(errors_for_this_case[~nan_mask])
            median_absolute_deviation = np.median([abs(x - median_val) for x in errors_for_this_case[~nan_mask]])

            best_val_for_case = min(errors_for_this_case[~nan_mask])
            max_val_to_survive = best_val_for_case + median_absolute_deviation
            new_candidates = []
            for x in candidates:
                if not np.isnan(x.ptscores[cases[0]]) and x.ptscores[cases[0]] <= max_val_to_survive:
                    new_candidates.append(x)
            candidates = new_candidates
            cases.pop(0)

        selected_individuals.append(random.choice(candidates))

    return selected_individuals

selBalAccLexicase(individuals, k)

Returns an individual that does the best on the fitness cases when considered one at a time in random order. http://faculty.hampshire.edu/lspector/pubs/lexicase-IEEE-TEC.pdf Matches will have 0 for a result while mismatches are 1 nan also causes a candidate to drop out

Parameters:

Name Type Description Default
individuals list

A list of individuals to select from.

 required
k int

The number of individuals to select.

 required

Returns:

Type Description
list

A list of selected individuals.
Source code in src/athenage/grape/grape.py 
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
def selBalAccLexicase(individuals: list, k: int) -> list:
    """Returns an individual that does the best on the fitness cases when
    considered one at a time in random order.
    http://faculty.hampshire.edu/lspector/pubs/lexicase-IEEE-TEC.pdf
    Matches will have 0 for a result while mismatches are 1
    nan also causes a candidate to drop out

    Args:
        individuals: A list of individuals to select from.
        k: The number of individuals to select.

    Returns: 
        A list of selected individuals.

    """

    selected_individuals = []

    nscores =  len(individuals[0].ptscores)

    for i in range(k):
        candidates = individuals
        cases = list(range(nscores))
        random.shuffle(cases)

        while len(cases) > 0 and len(candidates) > 1:
            newcandidates = [x for x in candidates if x.ptscores[cases[0]] == 0]
            if not newcandidates:
                break
            candidates = newcandidates
            cases.pop(0)

        selected_individuals.append(random.choice(candidates))

    return selected_individuals

selTournRoulette(individuals, k, fit_attr='fitness')

Pick two individuals from the population using the RouletteWheel selection method. Then return the better of the two individuals. Repeat to get k individuals to return. Matches the ATHENA tournament selection method

Parameters:

Name Type Description Default
individuals list

A list of individuals to select from.

 required
k int

The number of individuals to select.

 required
fit_attr str

The attribute of individuals to use as selection criterion

 'fitness'

Returns:

Type Description
list

A list containing the k selected individuals

Return:

Source code in src/athenage/grape/grape.py 
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
def selTournRoulette(individuals: list, k:int, fit_attr:str="fitness") -> list:
    """Pick two individuals from the population using the RouletteWheel selection
    method.  Then return the better of the two individuals. Repeat to get k individuals
    to return. Matches the ATHENA tournament selection method

    Args:
        individuals: A list of individuals to select from.
        k: The number of individuals to select.
        fit_attr: The attribute of individuals to use as selection criterion

    Returns: 
        A list containing the k selected individuals

    Return:

    """
    # sort the population in ascending order
    individuals = sorted(individuals, key=attrgetter(fit_attr), reverse=True)

    psum=[]
    n = len(individuals)

    # set the probabilities for each individual to be selected
    # if all identical then make all probabilities equal
    if individuals[0].fitness.values[0] == individuals[-1].fitness.values[0]:
        psum = [i.fitness/float(n) for i in individuals]
    else:
        maxfit = individuals[-1].fitness.values[0]
        minfit = individuals[0].fitness.values[0]
        # set first probability
        psum.append(-individuals[0].fitness.values[0] + maxfit + minfit)

        for i,ind in enumerate(individuals[1:], start=1):
            psum.append(-ind.fitness.values[0] + minfit + maxfit + psum[i-1])

        for i in range(n):
            psum[i] /= psum[-1]

    # select individuals
    chosen=[]
    for i in range(k):
        #use a binary search 
        aspirants=[]
        for j in range(2):
            cutoff = random.random()
            lowerbound = 0
            upperbound = n-1

            while(upperbound >= lowerbound):
                indindex=lowerbound + (upperbound-lowerbound)//2
                if psum[indindex] > cutoff:
                    upperbound = indindex-1
                else:
                    lowerbound = indindex+1

            lowerbound = min(n-1, lowerbound) 
            lowerbound = max(0, lowerbound)

            aspirants.append(individuals[lowerbound])

        selected = (min(aspirants, key=attrgetter(fit_attr)))

        chosen.append(selected)    

    return chosen

selTournamentWithoutInvalids(individuals, k, tournsize, fit_attr='fitness')

A simple tournament selection which avoids invalid individuals.

Parameters:

Name Type Description Default
individuals list

population to select from

 required
k int

number to select

 required
tournsize int

tournament size

 required
fit_attr str

attribute to use in selecting individuals

 'fitness'

Returns:

Type Description
list

list of selected individuals
Source code in src/athenage/grape/grape.py 
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
def selTournamentWithoutInvalids(individuals: list, k: int, tournsize: int, 
                                 fit_attr: str="fitness") -> list:
    """
    A simple tournament selection which avoids invalid individuals.

    Args:
        individuals: population to select from
        k: number to select
        tournsize: tournament size
        fit_attr: attribute to use in selecting individuals

    Returns:
        list of selected individuals
    """

    chosen = []
    valid_individuals = [i for i in individuals if not i.invalid]
    while len(chosen) < k:
        aspirants = random.sample(valid_individuals, tournsize)
        chosen.append(max(aspirants, key=attrgetter(fit_attr)))
    return chosen

sensible_initialization(ind_class, pop_size, bnf_grammar, min_init_depth, max_init_depth, codon_size, codon_consumption, genome_representation, genome_type)

Sensible initialization ensures valid individuals. Half the generated individuals are up to the depth of the max_init_depth and half are limited to that depth but may be smaller

Parameters:

Name Type Description Default
ind_class MetaCreator

Deap creator

 required
pop_size int

number of individuals to create

 required
bnf_grammar Grammar

BNF grammar

required
min_init_depth int

minimum depth for tree in mapping genome

 required
max_init_depth int

maximum depth for tree in mapping genome

 required
codon_size int

maximum to be stored in codon

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required
genome_representation str

either list or numpy

 required
genome_type str

standard/mcge/leap

 required

Returns:

Name Type Description
population list

list of individuals
Source code in src/athenage/grape/grape.py 
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
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
def sensible_initialization(ind_class: 'deap.creator.MetaCreator', pop_size:int, 
                          bnf_grammar:'grape.grape.Grammar', 
                          min_init_depth:int, 
                          max_init_depth:int, codon_size:int, codon_consumption:str,
                          genome_representation:str, genome_type:str) -> list:
    """Sensible initialization ensures valid individuals. Half the generated individuals
    are up to the depth of the max_init_depth and half are limited to that depth but may be
    smaller

    Args:
        ind_class: Deap creator
        pop_size: number of individuals to create
        bnf_grammar: BNF grammar 
        min_init_depth: minimum depth for tree in mapping genome
        max_init_depth: maximum depth for tree in mapping genome
        codon_size: maximum to be stored in codon
        codon_consumption: type of consumption when mapping (ie. lazy or eager)
        genome_representation: either list or numpy
        genome_type: standard/mcge/leap


    Returns:
        population: list of individuals
    """ 
    #Calculate the number of individuals to be generated with each method
    is_odd = pop_size % 2
    n_grow = int(pop_size/2)

    n_sets_grow = max_init_depth - min_init_depth + 1
    set_size = int(n_grow/n_sets_grow)
    remaining = n_grow % n_sets_grow

    n_full = n_grow + is_odd + remaining #if pop_size is odd, generate an extra ind with "full"

    #TODO check if it is possible to generate inds with max_init_depth        

    population = []
    #Generate inds using "Grow"
    for i in range(n_sets_grow):
        max_init_depth_ = min_init_depth + i
        for j in range(set_size):
            remainders = [] #it will register the choices
            possible_choices = [] #it will register the respective possible choices
            rules_used = [] #registers which rule used for setting the frame

            phenotype = bnf_grammar.start_rule
            remaining_NTs = ['<' + term + '>' for term in re.findall(r"\<([\(\)\w,-.]+)\>",phenotype)] #
            depths = [1]*len(remaining_NTs) #it keeps the depth of each branch
            idx_branch = 0 #index of the current branch being grown
            while len(remaining_NTs) != 0:
                idx_NT = bnf_grammar.non_terminals.index(remaining_NTs[0])
                total_options = [PR for PR in bnf_grammar.production_rules[idx_NT]]
                actual_options = [PR for PR in bnf_grammar.production_rules[idx_NT] if PR[5] + depths[idx_branch] <= max_init_depth_]
                Ch = random.choice(actual_options)
                phenotype = phenotype.replace(remaining_NTs[0], Ch[0], 1)
                depths[idx_branch] += 1
                if codon_consumption == 'eager':
                    remainders.append(Ch[3])
                    possible_choices.append(len(total_options))
                    rules_used.append(idx_NT)
                elif codon_consumption == 'lazy':
                    if len(total_options) > 1:
                        remainders.append(Ch[3])
                        possible_choices.append(len(total_options))
                        rules_used.append(idx_NT)

                if Ch[2] > 1:
                    if idx_branch == 0:
                        depths = [depths[idx_branch],]*Ch[2] + depths[idx_branch+1:]
                    else:
                        depths = depths[0:idx_branch] + [depths[idx_branch],]*Ch[2] + depths[idx_branch+1:]
                if Ch[1] == 'terminal':
                    idx_branch += 1

                remaining_NTs = ['<' + term + '>' for term in re.findall(r"\<([\(\)\w,-.]+)\>",phenotype)]

            #Generate the genome
            #remainders contain the choice for the rule
            #possible_choices simply tells you how many choices for that codon -- needed to generate actual codon value
            genome = create_genome(genome_type, bnf_grammar.nt_rule_size, codon_size)
            # genome = []
            if codon_consumption == 'eager' or codon_consumption == 'lazy':
                for k in range(len(remainders)):
                    codon = (random.randint(0,int(1e10)) % math.floor(((codon_size + 1) / possible_choices[k])) * possible_choices[k]) + remainders[k]
                    genome.add_codon(codon, rules_used[k])
                genome.finalize()
            else:
                raise ValueError("Unknown mapper")

            #Include a tail with 50% of the genome's size
            size_tail = int(0.5*genome.size())
            genome.add_tail(size_tail)

            #Initialize the individual and include in the population
            ind = ind_class(genome, bnf_grammar, max_init_depth_, codon_consumption)

            #Check if the individual was mapped correctly
            if remainders != ind.structure or phenotype != ind.phenotype or max(depths) != ind.depth:
                raise Exception('error in the mapping')
            population.append(ind)

    for i in range(n_full):
        remainders = [] #it will register the choices
        possible_choices = [] #it will register the respective possible choices
        rules_used = [] #registers which rule used for setting the frame

        phenotype = bnf_grammar.start_rule
        remaining_NTs = ['<' + term + '>' for term in re.findall(r"\<([\(\)\w,-.]+)\>",phenotype)] #
        depths = [1]*len(remaining_NTs) #it keeps the depth of each branch
        idx_branch = 0 #index of the current branch being grown

        while len(remaining_NTs) != 0:
            idx_NT = bnf_grammar.non_terminals.index(remaining_NTs[0])
            total_options = [PR for PR in bnf_grammar.production_rules[idx_NT]]
            actual_options = [PR for PR in bnf_grammar.production_rules[idx_NT] if PR[5] + depths[idx_branch] <= max_init_depth]
            recursive_options = [PR for PR in actual_options if PR[4]]
            if len(recursive_options) > 0:
                Ch = random.choice(recursive_options)
            else:
                Ch = random.choice(actual_options)
            phenotype = phenotype.replace(remaining_NTs[0], Ch[0], 1)
            depths[idx_branch] += 1
            if codon_consumption == 'eager':
                remainders.append(Ch[3])
                possible_choices.append(len(total_options))
                rules_used.append(idx_NT)
            elif codon_consumption == 'lazy':
                if len(total_options) > 1:
                    remainders.append(Ch[3])
                    possible_choices.append(len(total_options))
                    rules_used.append(idx_NT)

            if Ch[2] > 1:
                if idx_branch == 0:
                    depths = [depths[idx_branch],]*Ch[2] + depths[idx_branch+1:]
                else:
                    depths = depths[0:idx_branch] + [depths[idx_branch],]*Ch[2] + depths[idx_branch+1:]
            if Ch[1] == 'terminal':
                idx_branch += 1

            remaining_NTs = ['<' + term + '>' for term in re.findall(r"\<([\(\)\w,-.]+)\>",phenotype)]


        genome = create_genome(genome_type, bnf_grammar.nt_rule_size, codon_size)
        # genome = []
        if codon_consumption == 'eager' or codon_consumption == 'lazy':
            for k in range(len(remainders)):
                codon = (random.randint(0,int(1e10)) % math.floor(((codon_size + 1) / possible_choices[k])) * possible_choices[k]) + remainders[k]
                genome.add_codon(codon, rules_used[k])
            genome.finalize()
        else:
            raise ValueError("Unknown mapper")

        #Include a tail with 50% of the genome's size
        size_tail = int(0.5*genome.size())
        genome.add_tail(size_tail)

        #Initialise the individual and include in the population
        ind = ind_class(genome, bnf_grammar, max_init_depth, codon_consumption)

        #Check if the individual was mapped correctly
        if remainders != ind.structure or phenotype != ind.phenotype or max(depths) != ind.depth:
            raise Exception('error in the mapping')

        population.append(ind)    

    if genome_representation == 'list':
        return population
    elif genome_representation == 'numpy':
        for ind in population:
            ind.genome = np.array(ind.genome)
        return population
    else:
        raise ValueError("Unkonwn genome representation")

submap(ind, genome, bnf_grammar, start_codon, start_rule, codon_consumption)

Maps GE genome in individual to produce new phenotype

Parameters:

Name Type Description Default
ind Individual

individual to mutate

 required
bnf_grammar Grammar

BNF grammar

required
start_codon int

Index of starting codon for the subtree

 required
start_rule int

Index of rule for starting codon

 required
codon_consumption str

type of consumption when mapping (ie. lazy or eager)

 required

Returns:

Name Type Description
pos_end int

index of ending position in genome for the subtree
Source code in src/athenage/grape/grape.py 
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
def submap(ind: 'deap.creator.Individual', genome: list, 
          bnf_grammar:'grape.grape.Grammar', start_codon: int,
          start_rule:int, codon_consumption:str) -> int:
    """Maps GE genome in individual to produce new phenotype

    Args:
        ind: individual to mutate
        bnf_grammar: BNF grammar 
        start_codon: Index of starting codon for the subtree
        start_rule: Index of rule for starting codon
        codon_consumption: type of consumption when mapping (ie. lazy or eager)

    Returns:
        pos_end: index of ending position in genome for the subtree
    """ 

    # set initial ending to be rest of genome
    pos_end = ind.genome.total_codons()-1
    ind.genome = genome
    if codon_consumption == 'lazy':
        pos_end= submap_lazy(genome, bnf_grammar, start_codon, start_rule)
    elif codon_consumption == 'eager':
        pos_end = submap_eager(genome, bnf_grammar, start_codon, start_rule)
    else:
        raise ValueError("Unknown mapper")    

    return pos_end

submap_eager(genome, grammar, start_codon, start_rule)

Maps GE genome. Uses eager mapping so that every nonterminal in the grammar consumes a codon in the genome

Parameters:

Name Type Description Default
genome Genome

list of codons in genome

 required
grammar grammar

BNF grammar used to create individual phenotype

 required
start_codon int

Index in genome of starting codon

 required
start_rule int

Index of nonterminal in grammar

 required

Returns:

Name Type Description
codon_end tuple[str, int, int, int, bool, list]

ending codon for subtree of the expression tree
Source code in src/athenage/grape/grape.py 
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
def submap_eager(genome: "grape.grape.Genome", grammar: "grape.grape.grammar",
                  start_codon: int, start_rule: int) -> tuple[str, int, int, int, bool, list]:
    """Maps GE  genome. Uses eager mapping so that 
    every nonterminal in the grammar consumes a codon in the genome

    Args:
        genome: list of codons in genome
        grammar: BNF grammar used to create individual phenotype
        start_codon: Index in genome of starting codon
        start_rule: Index of nonterminal in grammar


    Returns:
        codon_end: ending codon for subtree of the expression tree
    """

    genome.reset_map_index()
    start_codon = start_codon[0]
    n_codons_used=0
    genome.next_read = start_codon
    next_NT = grammar.non_terminals[start_rule]
    phenotype = next_NT

    # while next_NT and idx_genome < len(genome):
    while next_NT:
        NT_index = grammar.non_terminals.index(next_NT)
        if genome.consumed(NT_index):
            break
        index_production_chosen = genome.get_next_codon(NT_index) % grammar.n_rules[NT_index]
        n_codons_used += 1
        phenotype = phenotype.replace(next_NT, grammar.production_rules[NT_index][index_production_chosen][0], 1)

        next_ = re.search(r"\<(\w+)\>",phenotype)
        if next_:
            next_NT = next_.group()
        else:
            next_NT = None

    if next_NT:
        invalid = True
    else:
        invalid = False

    return start_codon + n_codons_used -1

submap_lazy(genome, grammar, start_codon, start_rule)

Uses lazy mapping so that only nonterminals with more than one option in the grammar consume a codon in the genome

Parameters:

Name Type Description Default
genome Genome

list of codons in genome

 required
grammar grammar

BNF grammar used to create individual phenotype

 required
start_codon int

Index in genome of starting codon

 required
start_rule int

Index of nonterminal in grammar

 required

Returns:

Name Type Description
codon_end tuple[str, int, int, int, bool, list]

ending codon for subtree of the expression tree
Source code in src/athenage/grape/grape.py 
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
def submap_lazy(genome: "grape.grape.Genome", grammar: "grape.grape.grammar",
                  start_codon: int, start_rule: int) -> tuple[str, int, int, int, bool, list]:
    """ Uses lazy mapping so that 
    only nonterminals with more than one option in the grammar 
    consume a codon in the genome

    Args:
        genome: list of codons in genome
        grammar: BNF grammar used to create individual phenotype
        start_codon: Index in genome of starting codon
        start_rule: Index of nonterminal in grammar


    Returns:
        codon_end: ending codon for subtree of the expression tree
    """  

    genome.reset_map_index()
    start_codon = start_codon[0]
    n_codons_used=0
    genome.next_read = start_codon
    next_NT = grammar.non_terminals[start_rule]
    phenotype = next_NT

    while next_NT:
        NT_index = grammar.non_terminals.index(next_NT)
        if genome.consumed(NT_index):
            break        
        if grammar.n_rules[NT_index] == 1: #there is a single PR for this non-terminal
            index_production_chosen = 0        
        else: #we consume one codon, and add the index to the structure
            index_production_chosen = genome.get_next_codon(NT_index) % grammar.n_rules[NT_index]
            n_codons_used += 1

        phenotype = phenotype.replace(next_NT, grammar.production_rules[NT_index][index_production_chosen][0], 1)

        next_ = re.search(r"\<(\w+)\>",phenotype)
        if next_:
            next_NT = next_.group()
        else:
            next_NT = None

    if next_NT:
        invalid = True
    else:
        invalid = False

    return start_codon + n_codons_used -1