跳到内容

模板搜索空间

此文件是 TPOT 库的一部分。

TPOT 的当前版本由 Cedars-Sinai 计算生物医学系开发，开发人员包括：- Pedro Henrique Ribeiro (https://github.com/perib, https://www.linkedin.com/in/pedro-ribeiro/) - Anil Saini (anil.saini@cshs.org) - Jose Hernandez (jgh9094@gmail.com) - Jay Moran (jay.moran@cshs.org) - Nicholas Matsumoto (nicholas.matsumoto@cshs.org) - Hyunjun Choi (hyunjun.choi@cshs.org) - Gabriel Ketron (gabriel.ketron@cshs.org) - Miguel E. Hernandez (miguel.e.hernandez@cshs.org) - Jason Moore (moorejh28@gmail.com)

TPOT 的原始版本主要由宾夕法尼亚大学开发，开发人员包括：- Randal S. Olson (rso@randalolson.com) - Weixuan Fu (weixuanf@upenn.edu) - Daniel Angell (dpa34@drexel.edu) - Jason Moore (moorejh28@gmail.com) - 以及许多慷慨的开源贡献者

TPOT 是自由软件：您可以根据自由软件基金会发布的 GNU 宽松通用公共许可证（版本 3 或您选择的任何更高版本）的条款重新分发和/或修改它。

分发 TPOT 是希望它会有用，但没有任何担保；甚至不包含对适销性或特定用途适用性的默示担保。有关更多详细信息，请参阅 GNU 宽松通用公共许可证。

您应该已经收到 TPOT 附带的 GNU 宽松通用公共许可证副本。如果未收到，请参阅 https://gnu.ac.cn/licenses/。

DynamicUnionPipeline

基类：SearchSpace

源代码位于 tpot/search_spaces/pipelines/dynamicunion.py 
187
188
189
190
191
192
193
194
195
196
197
198
199
class DynamicUnionPipeline(SearchSpace):
    def __init__(self, search_space : SearchSpace, max_estimators=None, allow_repeats=False ) -> None:
        """
        Takes in a list of search spaces. will produce a pipeline of Sequential length. Each step in the pipeline will correspond to the the search space provided in the same index.
        """

        self.search_space = search_space
        self.max_estimators = max_estimators
        self.allow_repeats = allow_repeats

    def generate(self, rng=None):
        rng = np.random.default_rng(rng)
        return DynamicUnionPipelineIndividual(self.search_space, max_estimators=self.max_estimators, allow_repeats=self.allow_repeats, rng=rng)

__init__(search_space, max_estimators=None, allow_repeats=False)

接受一个搜索空间列表。将生成一个长度与列表相同的顺序管道。管道中的每个步骤将对应于同一索引中提供的搜索空间。

源代码位于 tpot/search_spaces/pipelines/dynamicunion.py 
188
189
190
191
192
193
194
195
def __init__(self, search_space : SearchSpace, max_estimators=None, allow_repeats=False ) -> None:
    """
    Takes in a list of search spaces. will produce a pipeline of Sequential length. Each step in the pipeline will correspond to the the search space provided in the same index.
    """

    self.search_space = search_space
    self.max_estimators = max_estimators
    self.allow_repeats = allow_repeats

DynamicUnionPipelineIndividual

基类：SklearnIndividual

接受一个搜索空间。将生成一个最多包含 max_estimators 个步骤的 FeatureUnion。FeatureUnion 的输出将是所有步骤的连接结果。

源代码位于 tpot/search_spaces/pipelines/dynamicunion.py 
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
class DynamicUnionPipelineIndividual(SklearnIndividual):
    """
    Takes in one search space.
    Will produce a FeatureUnion of up to max_estimators number of steps.
    The output of the FeatureUnion will the all of the steps concatenated together.

    """

    def __init__(self, search_space : SearchSpace, max_estimators=None, allow_repeats=False, rng=None) -> None:
        super().__init__()
        self.search_space = search_space

        if max_estimators is None:
            self.max_estimators = np.inf
        else:
            self.max_estimators = max_estimators

        self.allow_repeats = allow_repeats

        self.union_dict = {}

        if self.max_estimators == np.inf:
            init_max = 3
        else:
            init_max = self.max_estimators

        rng = np.random.default_rng(rng)

        for _ in range(rng.integers(1, init_max)):
            self._mutate_add_step(rng)


    def mutate(self, rng=None):
        rng = np.random.default_rng(rng)
        mutation_funcs = [self._mutate_add_step, self._mutate_remove_step, self._mutate_replace_step, self._mutate_note]
        rng.shuffle(mutation_funcs)
        for mutation_func in mutation_funcs:
            if mutation_func(rng):
                return True

    def _mutate_add_step(self, rng):
        rng = np.random.default_rng(rng)
        max_attempts = 10
        if len(self.union_dict) < self.max_estimators:
            for _ in range(max_attempts):
                new_step = self.search_space.generate(rng)
                if new_step.unique_id() not in self.union_dict:
                    self.union_dict[new_step.unique_id()] = new_step
                    return True
        return False

    def _mutate_remove_step(self, rng):
        rng = np.random.default_rng(rng)
        if len(self.union_dict) > 1:
            self.union_dict.pop( rng.choice(list(self.union_dict.keys())))  
            return True
        return False

    def _mutate_replace_step(self, rng):
        rng = np.random.default_rng(rng)        
        changed = self._mutate_remove_step(rng) or self._mutate_add_step(rng)
        return changed

    #TODO mutate one step or multiple?
    def _mutate_note(self, rng):
        rng = np.random.default_rng(rng)
        changed = False
        values = list(self.union_dict.values())
        for step in values:
            if rng.random() < 0.5:
                changed = step.mutate(rng) or changed

        self.union_dict = {step.unique_id(): step for step in values}

        return changed


    def crossover(self, other, rng=None):
        rng = np.random.default_rng(rng)

        cx_funcs = [self._crossover_swap_multiple_nodes, self._crossover_node]
        rng.shuffle(cx_funcs)
        for cx_func in cx_funcs:
            if cx_func(other, rng):
                return True

        return False


    def _crossover_swap_multiple_nodes(self, other, rng):
        rng = np.random.default_rng(rng)
        self_values = list(self.union_dict.values())
        other_values = list(other.union_dict.values())

        rng.shuffle(self_values)
        rng.shuffle(other_values)

        self_idx = rng.integers(0,len(self_values))
        other_idx = rng.integers(0,len(other_values))

        #Note that this is not one-point-crossover since the sequence doesn't matter. this is just a quick way to swap multiple random items
        self_values[:self_idx], other_values[:other_idx] = other_values[:other_idx], self_values[:self_idx]

        self.union_dict = {step.unique_id(): step for step in self_values}
        other.union_dict = {step.unique_id(): step for step in other_values}

        return True


    def _crossover_node(self, other, rng):
        rng = np.random.default_rng(rng)

        changed = False
        self_values = list(self.union_dict.values())
        other_values = list(other.union_dict.values())

        rng.shuffle(self_values)
        rng.shuffle(other_values)

        for self_step, other_step in zip(self_values, other_values):
            if rng.random() < 0.5:
                changed = self_step.crossover(other_step, rng) or changed

        self.union_dict = {step.unique_id(): step for step in self_values}
        other.union_dict = {step.unique_id(): step for step in other_values}

        return changed

    def export_pipeline(self, **kwargs):
        values = list(self.union_dict.values())
        return sklearn.pipeline.make_union(*[step.export_pipeline(**kwargs) for step in values])

    def unique_id(self):
        values = list(self.union_dict.values())
        l = [step.unique_id() for step in values]
        # if all items are strings, then sort them
        if all([isinstance(x, str) for x in l]):
            l.sort()
        l = ["FeatureUnion"] + l
        return TupleIndex(frozenset(l))

EstimatorNodeIndividual

基类：SklearnIndividual

请注意，ConfigurationSpace 不支持将 None 作为参数。作为替代，请使用特殊字符串 "". TPOT 将自动用 Python 的 None 替换此字符串的实例。

参数

名称 类型 描述 默认值
method type

要使用的评估器类

 必需的
space ConfigurationSpace | dict

要使用的超参数空间。如果传入 dict，则超参数是固定的，不进行学习。

 必需的
源代码位于 tpot/search_spaces/nodes/estimator_node.py 
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
class EstimatorNodeIndividual(SklearnIndividual):
    """
    Note that ConfigurationSpace does not support None as a parameter. Instead, use the special string "<NONE>". TPOT will automatically replace instances of this string with the Python None. 

    Parameters
    ----------
    method : type
        The class of the estimator to be used

    space : ConfigurationSpace|dict
        The hyperparameter space to be used. If a dict is passed, hyperparameters are fixed and not learned.

    """
    def __init__(self, method: type, 
                        space: ConfigurationSpace|dict, #TODO If a dict is passed, hyperparameters are fixed and not learned. Is this confusing? Should we make a second node type?
                        hyperparameter_parser: callable = None,
                        rng=None) -> None:
        super().__init__()
        self.method = method
        self.space = space

        if hyperparameter_parser is None:
            self.hyperparameter_parser = default_hyperparameter_parser
        else:
            self.hyperparameter_parser = hyperparameter_parser

        if isinstance(space, dict):
            self.hyperparameters = space
        else:
            rng = np.random.default_rng(rng)
            self.space.seed(rng.integers(0, 2**32))
            self.hyperparameters = dict(self.space.sample_configuration())

    def mutate(self, rng=None):
        if isinstance(self.space, dict): 
            return False

        rng = np.random.default_rng(rng)
        self.space.seed(rng.integers(0, 2**32))
        self.hyperparameters = dict(self.space.sample_configuration())
        return True

    def crossover(self, other, rng=None):
        if isinstance(self.space, dict):
            return False

        rng = np.random.default_rng(rng)
        if self.method != other.method:
            return False

        #loop through hyperparameters, randomly swap items in self.hyperparameters with items in other.hyperparameters
        for hyperparameter in self.space:
            if rng.choice([True, False]):
                if hyperparameter in other.hyperparameters:
                    self.hyperparameters[hyperparameter] = other.hyperparameters[hyperparameter]

        return True



    @final #this method should not be overridden, instead override hyperparameter_parser
    def export_pipeline(self, **kwargs):
        return self.method(**self.hyperparameter_parser(self.hyperparameters))

    def unique_id(self):
        #return a dictionary of the method and the hyperparameters
        method_str = self.method.__name__
        params = list(self.hyperparameters.keys())
        params = sorted(params)

        id_str = f"{method_str}({', '.join([f'{param}={self.hyperparameters[param]}' for param in params])})"

        return id_str

FSSIndividual

基类：SklearnIndividual

源代码位于 tpot/search_spaces/nodes/fss_node.py 
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
class FSSIndividual(SklearnIndividual):
    def __init__(   self,
                    subsets,
                    rng=None,
                ):

        """
        An individual for representing a specific FeatureSetSelector. 
        The FeatureSetSelector selects a feature list of list of predefined feature subsets.

        This instance will select one set initially. Mutation and crossover can swap the selected subset with another.

        Parameters
        ----------
        subsets : str or list, default=None
            Sets the subsets that the FeatureSetSeletor will select from if set as an option in one of the configuration dictionaries. 
            Features are defined by column names if using a Pandas data frame, or ints corresponding to indexes if using numpy arrays.
            - str : If a string, it is assumed to be a path to a csv file with the subsets. 
                The first column is assumed to be the name of the subset and the remaining columns are the features in the subset.
            - list or np.ndarray : If a list or np.ndarray, it is assumed to be a list of subsets (i.e a list of lists).
            - dict : A dictionary where keys are the names of the subsets and the values are the list of features.
            - int : If an int, it is assumed to be the number of subsets to generate. Each subset will contain one feature.
            - None : If None, each column will be treated as a subset. One column will be selected per subset.
        rng : int, np.random.Generator, optional
            The random number generator. The default is None.
            Only used to select the first subset.

        Returns
        -------
        None    
        """

        subsets = subsets
        rng = np.random.default_rng(rng)

        if isinstance(subsets, str):
            df = pd.read_csv(subsets,header=None,index_col=0)
            df['features'] = df.apply(lambda x: list([x[c] for c in df.columns]),axis=1)
            self.subset_dict = {}
            for row in df.index:
                self.subset_dict[row] = df.loc[row]['features']
        elif isinstance(subsets, dict):
            self.subset_dict = subsets
        elif isinstance(subsets, list) or isinstance(subsets, np.ndarray):
            self.subset_dict = {str(i):subsets[i] for i in range(len(subsets))}
        elif isinstance(subsets, int):
            self.subset_dict = {"{0}".format(i):i for i in range(subsets)}
        else:
            raise ValueError("Subsets must be a string, dictionary, list, int, or numpy array")

        self.names_list = list(self.subset_dict.keys())


        self.selected_subset_name = rng.choice(self.names_list)
        self.sel_subset = self.subset_dict[self.selected_subset_name]


    def mutate(self, rng=None):
        rng = np.random.default_rng(rng)
        #get list of names not including the current one
        names = [name for name in self.names_list if name != self.selected_subset_name]
        self.selected_subset_name = rng.choice(names)
        self.sel_subset = self.subset_dict[self.selected_subset_name]


    def crossover(self, other, rng=None):
        self.selected_subset_name = other.selected_subset_name
        self.sel_subset = other.sel_subset

    def export_pipeline(self, **kwargs):
        return FeatureSetSelector(sel_subset=self.sel_subset, name=self.selected_subset_name)


    def unique_id(self):
        id_str = "FeatureSetSelector({0})".format(self.selected_subset_name)
        return id_str

__init__(subsets, rng=None)

表示特定 FeatureSetSelector 的个体。FeatureSetSelector 从预定义的特征子集列表中选择一个特征列表。

此实例最初将选择一个集合。突变和交叉操作可以将其选定的子集与另一个子集进行交换。

参数

名称 类型 描述 默认值
subsets str 或 list

设置 FeatureSetSelector 将从中选择的子集（如果在配置字典中设置为选项）。如果使用 Pandas 数据帧，特征由列名定义；如果使用 numpy 数组，则由对应索引的整数定义。- str：如果为字符串，则假定它是包含子集的 csv 文件路径。第一列被假定为子集名称，其余列为子集中的特征。- list 或 np.ndarray：如果为 list 或 np.ndarray，则假定它是一个子集列表（即一个列表的列表）。- dict：一个字典，其中键是子集的名称，值是特征列表。- int：如果为 int，则假定是要生成的子集数量。每个子集将包含一个特征。- None：如果为 None，则每列将被视为一个子集。每个子集将选择一列。

 None
rng (int, Generator)

随机数生成器。默认值为 None。仅用于选择第一个子集。

 None

返回

类型 描述
None
源代码位于 tpot/search_spaces/nodes/fss_node.py 
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
def __init__(   self,
                subsets,
                rng=None,
            ):

    """
    An individual for representing a specific FeatureSetSelector. 
    The FeatureSetSelector selects a feature list of list of predefined feature subsets.

    This instance will select one set initially. Mutation and crossover can swap the selected subset with another.

    Parameters
    ----------
    subsets : str or list, default=None
        Sets the subsets that the FeatureSetSeletor will select from if set as an option in one of the configuration dictionaries. 
        Features are defined by column names if using a Pandas data frame, or ints corresponding to indexes if using numpy arrays.
        - str : If a string, it is assumed to be a path to a csv file with the subsets. 
            The first column is assumed to be the name of the subset and the remaining columns are the features in the subset.
        - list or np.ndarray : If a list or np.ndarray, it is assumed to be a list of subsets (i.e a list of lists).
        - dict : A dictionary where keys are the names of the subsets and the values are the list of features.
        - int : If an int, it is assumed to be the number of subsets to generate. Each subset will contain one feature.
        - None : If None, each column will be treated as a subset. One column will be selected per subset.
    rng : int, np.random.Generator, optional
        The random number generator. The default is None.
        Only used to select the first subset.

    Returns
    -------
    None    
    """

    subsets = subsets
    rng = np.random.default_rng(rng)

    if isinstance(subsets, str):
        df = pd.read_csv(subsets,header=None,index_col=0)
        df['features'] = df.apply(lambda x: list([x[c] for c in df.columns]),axis=1)
        self.subset_dict = {}
        for row in df.index:
            self.subset_dict[row] = df.loc[row]['features']
    elif isinstance(subsets, dict):
        self.subset_dict = subsets
    elif isinstance(subsets, list) or isinstance(subsets, np.ndarray):
        self.subset_dict = {str(i):subsets[i] for i in range(len(subsets))}
    elif isinstance(subsets, int):
        self.subset_dict = {"{0}".format(i):i for i in range(subsets)}
    else:
        raise ValueError("Subsets must be a string, dictionary, list, int, or numpy array")

    self.names_list = list(self.subset_dict.keys())


    self.selected_subset_name = rng.choice(self.names_list)
    self.sel_subset = self.subset_dict[self.selected_subset_name]

FSSNode

基类：SearchSpace

源代码位于 tpot/search_spaces/nodes/fss_node.py 
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
class FSSNode(SearchSpace):
    def __init__(self,                     
                    subsets,
                ):
        """
        A search space for a FeatureSetSelector. 
        The FeatureSetSelector selects a feature list of list of predefined feature subsets.

        Parameters
        ----------
        subsets : str or list, default=None
            Sets the subsets that the FeatureSetSeletor will select from if set as an option in one of the configuration dictionaries. 
            Features are defined by column names if using a Pandas data frame, or ints corresponding to indexes if using numpy arrays.
            - str : If a string, it is assumed to be a path to a csv file with the subsets. 
                The first column is assumed to be the name of the subset and the remaining columns are the features in the subset.
            - list or np.ndarray : If a list or np.ndarray, it is assumed to be a list of subsets (i.e a list of lists).
            - dict : A dictionary where keys are the names of the subsets and the values are the list of features.
            - int : If an int, it is assumed to be the number of subsets to generate. Each subset will contain one feature.
            - None : If None, each column will be treated as a subset. One column will be selected per subset.

        Returns
        -------
        None    

        """

        self.subsets = subsets

    def generate(self, rng=None) -> SklearnIndividual:
        return FSSIndividual(   
            subsets=self.subsets,
            rng=rng,
            )

__init__(subsets)

FeatureSetSelector 的搜索空间。FeatureSetSelector 从预定义的特征子集列表中选择一个特征列表。

参数

名称 类型 描述 默认值
subsets str 或 list

设置 FeatureSetSelector 将从中选择的子集（如果在配置字典中设置为选项）。如果使用 Pandas 数据帧，特征由列名定义；如果使用 numpy 数组，则由对应索引的整数定义。- str：如果为字符串，则假定它是包含子集的 csv 文件路径。第一列被假定为子集名称，其余列为子集中的特征。- list 或 np.ndarray：如果为 list 或 np.ndarray，则假定它是一个子集列表（即一个列表的列表）。- dict：一个字典，其中键是子集的名称，值是特征列表。- int：如果为 int，则假定是要生成的子集数量。每个子集将包含一个特征。- None：如果为 None，则每列将被视为一个子集。每个子集将选择一列。

 None

返回

类型 描述
None
源代码位于 tpot/search_spaces/nodes/fss_node.py 
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
def __init__(self,                     
                subsets,
            ):
    """
    A search space for a FeatureSetSelector. 
    The FeatureSetSelector selects a feature list of list of predefined feature subsets.

    Parameters
    ----------
    subsets : str or list, default=None
        Sets the subsets that the FeatureSetSeletor will select from if set as an option in one of the configuration dictionaries. 
        Features are defined by column names if using a Pandas data frame, or ints corresponding to indexes if using numpy arrays.
        - str : If a string, it is assumed to be a path to a csv file with the subsets. 
            The first column is assumed to be the name of the subset and the remaining columns are the features in the subset.
        - list or np.ndarray : If a list or np.ndarray, it is assumed to be a list of subsets (i.e a list of lists).
        - dict : A dictionary where keys are the names of the subsets and the values are the list of features.
        - int : If an int, it is assumed to be the number of subsets to generate. Each subset will contain one feature.
        - None : If None, each column will be treated as a subset. One column will be selected per subset.

    Returns
    -------
    None    

    """

    self.subsets = subsets

FeatureSetSelector

基类：BaseEstimator, SelectorMixin

选择预定义的特征子集。

源代码位于 tpot/builtin_modules/feature_set_selector.py 
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
class FeatureSetSelector(BaseEstimator, SelectorMixin):
    """
    Select predefined feature subsets.


    """

    def __init__(self, sel_subset=None, name=None):
        """Create a FeatureSetSelector object.

        Parameters
        ----------
        sel_subset: list or int
            If X is a dataframe, items in sel_subset list must correspond to column names
            If X is a numpy array, items in sel_subset list must correspond to column indexes
            int: index of a single column
        Returns
        -------
        None

        """
        self.name = name
        self.sel_subset = sel_subset


    def fit(self, X, y=None):
        """Fit FeatureSetSelector for feature selection

        Parameters
        ----------
        X: array-like of shape (n_samples, n_features)
            The training input samples.
        y: array-like, shape (n_samples,)
            The target values (integers that correspond to classes in classification, real numbers in regression).

        Returns
        -------
        self: object
            Returns a copy of the estimator
        """
        if isinstance(self.sel_subset, int) or isinstance(self.sel_subset, str):
            self.sel_subset = [self.sel_subset]

        #generate  self.feat_list_idx
        if isinstance(X, pd.DataFrame):
            self.feature_names_in_ = X.columns.tolist()
            self.feat_list_idx = sorted([self.feature_names_in_.index(feat) for feat in self.sel_subset])


        elif isinstance(X, np.ndarray):
            self.feature_names_in_ = None#list(range(X.shape[1]))

            self.feat_list_idx = sorted(self.sel_subset)

        n_features = X.shape[1]
        self.mask = np.zeros(n_features, dtype=bool)
        self.mask[np.asarray(self.feat_list_idx)] = True

        return self

    #TODO keep returned as dataframe if input is dataframe? may not be consistent with sklearn

    # def transform(self, X):

    def _get_tags(self):
        tags = {"allow_nan": True, "requires_y": False}
        return tags

    def _get_support_mask(self):
        """
        Get the boolean mask indicating which features are selected
        Returns
        -------
        support : boolean array of shape [# input features]
            An element is True iff its corresponding feature is selected for
            retention.
        """
        return self.mask

__init__(sel_subset=None, name=None)

创建一个 FeatureSetSelector 对象。

参数

名称 类型 描述 默认值
sel_subset

如果 X 是数据帧，sel_subset 列表中的项必须对应于列名 如果 X 是 numpy 数组，sel_subset 列表中的项必须对应于列索引 int: 单列索引

 None

返回

类型 描述
None
源代码位于 tpot/builtin_modules/feature_set_selector.py 
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
def __init__(self, sel_subset=None, name=None):
    """Create a FeatureSetSelector object.

    Parameters
    ----------
    sel_subset: list or int
        If X is a dataframe, items in sel_subset list must correspond to column names
        If X is a numpy array, items in sel_subset list must correspond to column indexes
        int: index of a single column
    Returns
    -------
    None

    """
    self.name = name
    self.sel_subset = sel_subset

fit(X, y=None)

拟合 FeatureSetSelector 进行特征选择

参数

名称 类型 描述 默认值
X

训练输入样本。

 必需的
y

目标值（分类中对应于类别的整数，回归中的实数）。

 None

返回

名称 类型 描述
self 对象

返回评估器的副本
源代码位于 tpot/builtin_modules/feature_set_selector.py 
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
def fit(self, X, y=None):
    """Fit FeatureSetSelector for feature selection

    Parameters
    ----------
    X: array-like of shape (n_samples, n_features)
        The training input samples.
    y: array-like, shape (n_samples,)
        The target values (integers that correspond to classes in classification, real numbers in regression).

    Returns
    -------
    self: object
        Returns a copy of the estimator
    """
    if isinstance(self.sel_subset, int) or isinstance(self.sel_subset, str):
        self.sel_subset = [self.sel_subset]

    #generate  self.feat_list_idx
    if isinstance(X, pd.DataFrame):
        self.feature_names_in_ = X.columns.tolist()
        self.feat_list_idx = sorted([self.feature_names_in_.index(feat) for feat in self.sel_subset])


    elif isinstance(X, np.ndarray):
        self.feature_names_in_ = None#list(range(X.shape[1]))

        self.feat_list_idx = sorted(self.sel_subset)

    n_features = X.shape[1]
    self.mask = np.zeros(n_features, dtype=bool)
    self.mask[np.asarray(self.feat_list_idx)] = True

    return self

GeneticFeatureSelectorNode

基类：SearchSpace

源代码位于 tpot/search_spaces/nodes/genetic_feature_selection.py 
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
class GeneticFeatureSelectorNode(SearchSpace):
    def __init__(self,                     
                    n_features,
                    start_p=0.2,
                    mutation_rate = 0.1,
                    crossover_rate = 0.1,
                    mutation_rate_rate = 0, # These are still experimental but seem to help. Theory is that it takes slower steps as it gets closer to the optimal solution.
                    crossover_rate_rate = 0,# Otherwise is mutation_rate is too small, it takes forever, and if its too large, it never converges.
                    ):
        """
        A node that generates a GeneticFeatureSelectorIndividual. Uses genetic algorithm to select novel subsets of features.

        Parameters
        ----------
        n_features : int
            Number of features in the dataset.
        start_p : float
            Probability of selecting a given feature for the initial subset of features.
        mutation_rate : float
            Probability of adding/removing a feature from the subset of features.
        crossover_rate : float
            Probability of swapping a feature between two subsets of features.
        mutation_rate_rate : float
            Probability of changing the mutation rate. (experimental)
        crossover_rate_rate : float
            Probability of changing the crossover rate. (experimental)

        """

        self.n_features = n_features
        self.start_p = start_p
        self.mutation_rate = mutation_rate
        self.crossover_rate = crossover_rate
        self.mutation_rate_rate = mutation_rate_rate
        self.crossover_rate_rate = crossover_rate_rate


    def generate(self, rng=None) -> SklearnIndividual:
        return GeneticFeatureSelectorIndividual(   mask=self.n_features,
                                                    start_p=self.start_p,
                                                    mutation_rate=self.mutation_rate,
                                                    crossover_rate=self.crossover_rate,
                                                    mutation_rate_rate=self.mutation_rate_rate,
                                                    crossover_rate_rate=self.crossover_rate_rate,
                                                    rng=rng
                                                )

__init__(n_features, start_p=0.2, mutation_rate=0.1, crossover_rate=0.1, mutation_rate_rate=0, crossover_rate_rate=0)

生成 GeneticFeatureSelectorIndividual 的节点。使用遗传算法选择新的特征子集。

参数

名称 类型 描述 默认值
n_features int

数据集中的特征数量。

 必需的
start_p float

为初始特征子集选择给定特征的概率。

 0.2
mutation_rate float

从特征子集中添加/移除特征的概率。

 0.1
crossover_rate float

在两个特征子集之间交换特征的概率。

 0.1
mutation_rate_rate float

改变突变率的概率。（实验性）

 0
crossover_rate_rate float

改变交叉率的概率。（实验性）

 0
源代码位于 tpot/search_spaces/nodes/genetic_feature_selection.py 
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
def __init__(self,                     
                n_features,
                start_p=0.2,
                mutation_rate = 0.1,
                crossover_rate = 0.1,
                mutation_rate_rate = 0, # These are still experimental but seem to help. Theory is that it takes slower steps as it gets closer to the optimal solution.
                crossover_rate_rate = 0,# Otherwise is mutation_rate is too small, it takes forever, and if its too large, it never converges.
                ):
    """
    A node that generates a GeneticFeatureSelectorIndividual. Uses genetic algorithm to select novel subsets of features.

    Parameters
    ----------
    n_features : int
        Number of features in the dataset.
    start_p : float
        Probability of selecting a given feature for the initial subset of features.
    mutation_rate : float
        Probability of adding/removing a feature from the subset of features.
    crossover_rate : float
        Probability of swapping a feature between two subsets of features.
    mutation_rate_rate : float
        Probability of changing the mutation rate. (experimental)
    crossover_rate_rate : float
        Probability of changing the crossover rate. (experimental)

    """

    self.n_features = n_features
    self.start_p = start_p
    self.mutation_rate = mutation_rate
    self.crossover_rate = crossover_rate
    self.mutation_rate_rate = mutation_rate_rate
    self.crossover_rate_rate = crossover_rate_rate

GraphKey

可用作图键的类。

参数

名称 类型 描述 默认值
graph 图

用作键的图。使用节点属性进行哈希计算。

 必需的
matched_label str

考虑用于哈希计算的节点属性。

 'label'
源代码位于 tpot/search_spaces/pipelines/graph.py 
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
class GraphKey():
    '''
    A class that can be used as a key for a graph.

    Parameters
    ----------
    graph : (nx.Graph)
        The graph to use as a key. Node Attributes are used for the hash.
    matched_label : (str)
        The node attribute to consider for the hash.
    '''

    def __init__(self, graph, matched_label='label') -> None:#['hyperparameters', 'method_class']) -> None:


        self.graph = graph
        self.matched_label = matched_label
        self.node_match = partial(node_match, matched_labels=[matched_label])
        self.key = int(nx.weisfeiler_lehman_graph_hash(self.graph, node_attr=self.matched_label),16) #hash(tuple(sorted([val for (node, val) in self.graph.degree()])))


    #If hash is different, node is definitely different
    # https://arxiv.org/pdf/2002.06653.pdf
    def __hash__(self) -> int:

        return self.key

    #If hash is same, use __eq__ to know if they are actually different
    def __eq__(self, other):
        return nx.is_isomorphic(self.graph, other.graph, node_match=self.node_match)

GraphPipelineIndividual

基类：SklearnIndividual

定义了有向无环图（DAG）形状的管道搜索空间。如果需要，可以单独定义根节点、叶节点和内部节点的搜索空间。每个图将有一个单一根节点作为最终评估器，该节点从 root_search_space 中抽取。如果定义了 leaf_search_space，管道中的所有叶节点将从该搜索空间中抽取。如果未定义 leaf_search_space，所有叶节点将从 inner_search_space 中抽取。非叶节点或根节点的节点将从 inner_search_space 中抽取。如果未定义 inner_search_space，将没有内部节点。

cross_val_predict_cv、method、memory 和 use_label_encoder 在导出管道时传递给 GraphPipeline 对象，并且不直接在搜索空间中使用。

导出为 GraphPipeline 对象。

参数

名称 类型 描述 默认值
root_search_space SearchSpace

图的根节点的搜索空间。此节点将是管道中的最终评估器。

 必需的
inner_search_space SearchSpace

图的内部节点的搜索空间。如果未定义，将没有内部节点。

 None
leaf_search_space SearchSpace

图的叶节点的搜索空间。如果未定义，叶节点将从 inner_search_space 中抽取。

 None
crossover_same_depth bool

如果为 True，交叉操作将仅发生在图中同一深度的节点之间。如果为 False，交叉操作将发生在任何深度的节点之间。

 False
cross_val_predict_cv Union[int, Callable]

确定内部分类器或回归器中使用的交叉验证分割策略

 0
method str

用于内部分类器或回归器的预测方法。如果为 'auto'，它将按 predict_proba、decision_function 或 predict 的顺序尝试使用。

 'auto'
memory

用于缓存节点的输入和输出，以防止重复拟合或计算量大的变换。默认情况下，不执行缓存。如果给定字符串，则为缓存目录的路径。

 必需的
use_label_encoder bool

如果为 True，使用标签编码器将标签编码为从 0 到 N。如果为 False，不使用标签编码器。主要用于需要标签为从 0 到 N 的整数的分类器（如 XGBoost）。也可以是 sklearn.preprocessing.LabelEncoder 对象。如果是，则使用该标签编码器。

 False
rng

采样第一个图实例的种子。

 None
源代码位于 tpot/search_spaces/pipelines/graph.py 
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
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
276
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
315
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
389
390
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
467
468
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
497
498
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
543
544
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
593
594
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
622
623
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
672
673
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
class GraphPipelineIndividual(SklearnIndividual):
    """
        Defines a search space of pipelines in the shape of a Directed Acyclic Graphs. The search spaces for root, leaf, and inner nodes can be defined separately if desired.
        Each graph will have a single root serving as the final estimator which is drawn from the `root_search_space`. If the `leaf_search_space` is defined, all leaves 
        in the pipeline will be drawn from that search space. If the `leaf_search_space` is not defined, all leaves will be drawn from the `inner_search_space`.
        Nodes that are not leaves or roots will be drawn from the `inner_search_space`. If the `inner_search_space` is not defined, there will be no inner nodes.

        `cross_val_predict_cv`, `method`, `memory`, and `use_label_encoder` are passed to the GraphPipeline object when the pipeline is exported and not directly used in the search space.

        Exports to a GraphPipeline object.

        Parameters
        ----------

        root_search_space: SearchSpace
            The search space for the root node of the graph. This node will be the final estimator in the pipeline.

        inner_search_space: SearchSpace, optional
            The search space for the inner nodes of the graph. If not defined, there will be no inner nodes.

        leaf_search_space: SearchSpace, optional
            The search space for the leaf nodes of the graph. If not defined, the leaf nodes will be drawn from the inner_search_space.

        crossover_same_depth: bool, optional
            If True, crossover will only occur between nodes at the same depth in the graph. If False, crossover will occur between nodes at any depth.

        cross_val_predict_cv: int, cross-validation generator or an iterable, optional
            Determines the cross-validation splitting strategy used in inner classifiers or regressors

        method: str, optional
            The prediction method to use for the inner classifiers or regressors. If 'auto', it will try to use predict_proba, decision_function, or predict in that order.

        memory: str or object with the joblib.Memory interface, optional
            Used to cache the input and outputs of nodes to prevent refitting or computationally heavy transformations. By default, no caching is performed. If a string is given, it is the path to the caching directory.

        use_label_encoder: bool, optional
            If True, the label encoder is used to encode the labels to be 0 to N. If False, the label encoder is not used.
            Mainly useful for classifiers (XGBoost) that require labels to be ints from 0 to N.
            Can also be a sklearn.preprocessing.LabelEncoder object. If so, that label encoder is used.

        rng: int, RandomState instance or None, optional
            Seed for sampling the first graph instance. 

        """

    def __init__(
            self,  
            root_search_space: SearchSpace, 
            leaf_search_space: SearchSpace = None, 
            inner_search_space: SearchSpace = None, 
            max_size: int = np.inf,
            crossover_same_depth: bool = False,
            cross_val_predict_cv: Union[int, Callable] = 0, #signature function(estimator, X, y=none)
            method: str = 'auto',
            use_label_encoder: bool = False,
            rng=None):

        super().__init__()

        self.__debug = False

        rng = np.random.default_rng(rng)

        self.root_search_space = root_search_space
        self.leaf_search_space = leaf_search_space
        self.inner_search_space = inner_search_space
        self.max_size = max_size
        self.crossover_same_depth = crossover_same_depth

        self.cross_val_predict_cv = cross_val_predict_cv
        self.method = method
        self.use_label_encoder = use_label_encoder

        self.root = self.root_search_space.generate(rng)
        self.graph = nx.DiGraph()
        self.graph.add_node(self.root)

        if self.leaf_search_space is not None:
            self.leaf = self.leaf_search_space.generate(rng)
            self.graph.add_node(self.leaf)
            self.graph.add_edge(self.root, self.leaf)

        if self.inner_search_space is None and self.leaf_search_space is None:
            self.mutate_methods_list = [self._mutate_node]
            self.crossover_methods_list = [self._crossover_swap_branch,]#[self._crossover_swap_branch, self._crossover_swap_node, self._crossover_take_branch]  #TODO self._crossover_nodes, 

        else:
            self.mutate_methods_list = [self._mutate_insert_leaf, self._mutate_insert_inner_node, self._mutate_remove_node, self._mutate_node, self._mutate_insert_bypass_node]
            self.crossover_methods_list = [self._crossover_swap_branch, self._crossover_nodes, self._crossover_take_branch ]#[self._crossover_swap_branch, self._crossover_swap_node, self._crossover_take_branch]  #TODO self._crossover_nodes, 

        self.merge_duplicated_nodes_toggle = True

        self.graphkey = None


    def mutate(self, rng=None):
        rng = np.random.default_rng(rng)
        rng.shuffle(self.mutate_methods_list)
        for mutate_method in self.mutate_methods_list:
            if mutate_method(rng=rng):

                if self.merge_duplicated_nodes_toggle:
                    self._merge_duplicated_nodes()

                if self.__debug:
                    print(mutate_method)

                    if self.root not in self.graph.nodes:
                        print('lost root something went wrong with ', mutate_method)

                    if len(self.graph.predecessors(self.root)) > 0:
                        print('root has parents ', mutate_method)

                    if any([n in nx.ancestors(self.graph,n) for n in self.graph.nodes]):
                        print('a node is connecting to itself...')

                    if self.__debug:
                        try:
                            nx.find_cycle(self.graph)
                            print('something went wrong with ', mutate_method)
                        except:
                            pass

                self.graphkey = None

        return False




    def _mutate_insert_leaf(self, rng=None):
        rng = np.random.default_rng(rng)
        if self.max_size > self.graph.number_of_nodes():
            sorted_nodes_list = list(self.graph.nodes)
            rng.shuffle(sorted_nodes_list) #TODO: sort by number of children and/or parents? bias model one way or another
            for node in sorted_nodes_list:
                #if leafs are protected, check if node is a leaf
                #if node is a leaf, skip because we don't want to add node on top of node
                if (self.leaf_search_space is not None #if leafs are protected
                    and   len(list(self.graph.successors(node))) == 0 #if node is leaf
                    and  len(list(self.graph.predecessors(node))) > 0 #except if node is root, in which case we want to add a leaf even if it happens to be a leaf too
                    ):

                    continue

                #If node *is* the root or is not a leaf, add leaf node. (dont want to add leaf on top of leaf)
                if self.leaf_search_space is not None:
                    new_node = self.leaf_search_space.generate(rng)
                else:
                    new_node = self.inner_search_space.generate(rng)

                self.graph.add_node(new_node)
                self.graph.add_edge(node, new_node)
                return True

        return False

    def _mutate_insert_inner_node(self, rng=None):
        """
        Finds an edge in the graph and inserts a new node between the two nodes. Removes the edge between the two nodes.
        """
        rng = np.random.default_rng(rng)
        if self.max_size > self.graph.number_of_nodes():
            sorted_nodes_list = list(self.graph.nodes)
            sorted_nodes_list2 = list(self.graph.nodes)
            rng.shuffle(sorted_nodes_list) #TODO: sort by number of children and/or parents? bias model one way or another
            rng.shuffle(sorted_nodes_list2)
            for node in sorted_nodes_list:
                #loop through children of node
                for child_node in list(self.graph.successors(node)):

                    if child_node is not node and child_node not in nx.ancestors(self.graph, node):
                        if self.leaf_search_space is not None:
                            #If if we are protecting leafs, dont add connection into a leaf
                            if len(list(nx.descendants(self.graph,node))) ==0 :
                                continue

                        new_node = self.inner_search_space.generate(rng)

                        self.graph.add_node(new_node)
                        self.graph.add_edges_from([(node, new_node), (new_node, child_node)])
                        self.graph.remove_edge(node, child_node)
                        return True

        return False


    def _mutate_remove_node(self, rng=None):
        '''
        Removes a randomly chosen node and connects its parents to its children.
        If the node is the only leaf for an inner node and 'leaf_search_space' is not none, we do not remove it.
        '''
        rng = np.random.default_rng(rng)
        nodes_list = list(self.graph.nodes)
        nodes_list.remove(self.root)
        leaves = get_leaves(self.graph)

        while len(nodes_list) > 0:
            node = rng.choice(nodes_list)
            nodes_list.remove(node)

            if self.leaf_search_space is not None and len(list(nx.descendants(self.graph,node))) == 0 : #if the node is a leaf
                if len(leaves) <= 1:
                    continue #dont remove the last leaf
                leaf_parents = self.graph.predecessors(node)

                # if any of the parents of the node has one one child, continue
                if any([len(list(self.graph.successors(lp))) < 2 for lp in leaf_parents]): #dont remove a leaf if it is the only input into another node.
                    continue

                remove_and_stitch(self.graph, node)
                remove_nodes_disconnected_from_node(self.graph, self.root)
                return True

            else:
                remove_and_stitch(self.graph, node)
                remove_nodes_disconnected_from_node(self.graph, self.root)
                return True

        return False



    def _mutate_node(self, rng=None):
        '''
        Mutates the hyperparameters for a randomly chosen node in the graph.
        '''
        rng = np.random.default_rng(rng)
        sorted_nodes_list = list(self.graph.nodes)
        rng.shuffle(sorted_nodes_list)
        completed_one = False
        for node in sorted_nodes_list:
            if node.mutate(rng):
                return True
        return False

    def _mutate_remove_edge(self, rng=None):
        '''
        Deletes an edge as long as deleting that edge does not make the graph disconnected.
        '''
        rng = np.random.default_rng(rng)
        sorted_nodes_list = list(self.graph.nodes)
        rng.shuffle(sorted_nodes_list)
        for child_node in sorted_nodes_list:
            parents = list(self.graph.predecessors(child_node))
            if len(parents) > 1: # if it has more than one parent, you can remove an edge (if this is the only child of a node, it will become a leaf)

                for parent_node in parents:
                    # if removing the egde will make the parent_node a leaf node, skip
                    if self.leaf_search_space is not None and len(list(self.graph.successors(parent_node))) < 2:
                        continue

                    self.graph.remove_edge(parent_node, child_node)
                    return True
        return False   

    def _mutate_add_edge(self, rng=None):
        '''
        Randomly add an edge from a node to another node that is not an ancestor of the first node.
        '''
        rng = np.random.default_rng(rng)
        sorted_nodes_list = list(self.graph.nodes)
        rng.shuffle(sorted_nodes_list)
        for child_node in sorted_nodes_list:
            for parent_node in sorted_nodes_list:
                if self.leaf_search_space is not None:
                    if len(list(self.graph.successors(parent_node))) == 0:
                        continue

                # skip if
                # - parent and child are the same node
                # - edge already exists
                # - child is an ancestor of parent
                if  (child_node is not parent_node) and not self.graph.has_edge(parent_node,child_node) and (child_node not in nx.ancestors(self.graph, parent_node)):
                    self.graph.add_edge(parent_node,child_node)
                    return True

        return False

    def _mutate_insert_bypass_node(self, rng=None):
        """
        Pick two nodes (doesn't necessarily need to be connected). Create a new node. connect one node to the new node and the new node to the other node.
        Does not remove any edges.
        """
        rng = np.random.default_rng(rng)
        if self.max_size > self.graph.number_of_nodes():
            sorted_nodes_list = list(self.graph.nodes)
            sorted_nodes_list2 = list(self.graph.nodes)
            rng.shuffle(sorted_nodes_list) #TODO: sort by number of children and/or parents? bias model one way or another
            rng.shuffle(sorted_nodes_list2)
            for node in sorted_nodes_list:
                for child_node in sorted_nodes_list2:
                    if child_node is not node and child_node not in nx.ancestors(self.graph, node):
                        if self.leaf_search_space is not None:
                            #If if we are protecting leafs, dont add connection into a leaf
                            if len(list(nx.descendants(self.graph,node))) ==0 :
                                continue

                        new_node = self.inner_search_space.generate(rng)

                        self.graph.add_node(new_node)
                        self.graph.add_edges_from([(node, new_node), (new_node, child_node)])
                        return True

        return False


    def crossover(self, ind2, rng=None):
        '''
        self is the first individual, ind2 is the second individual
        If crossover_same_depth, it will select graphindividuals at the same recursive depth.
        Otherwise, it will select graphindividuals randomly from the entire graph and its subgraphs.

        This does not impact graphs without subgraphs. And it does not impacts nodes that are not graphindividuals. Cros
        '''

        rng = np.random.default_rng(rng)

        rng.shuffle(self.crossover_methods_list)

        finished = False

        for crossover_method in self.crossover_methods_list:
            if crossover_method(ind2, rng=rng):
                self._merge_duplicated_nodes()
                finished = True
                break

        if self.__debug:
            try:
                nx.find_cycle(self.graph)
                print('something went wrong with ', crossover_method)
            except:
                pass

        if finished:
            self.graphkey = None

        return finished


    def _crossover_swap_branch(self, G2, rng=None):
        '''
        swaps a branch from parent1 with a branch from parent2. does not modify parent2
        '''
        rng = np.random.default_rng(rng)

        if self.crossover_same_depth:
            pair_gen = select_nodes_same_depth(self.graph, self.root, G2.graph, G2.root, rng=rng)
        else:
            pair_gen = select_nodes_randomly(self.graph, G2.graph, rng=rng)

        for node1, node2 in pair_gen:
            #TODO: if root is in inner_search_space, then do use it?
            if node1 is self.root or node2 is G2.root: #dont want to add root as inner node
                continue

            #check if node1 is a leaf and leafs are protected, don't add an input to the leave
            if self.leaf_search_space is not None: #if we are protecting leaves,
                node1_is_leaf = len(list(self.graph.successors(node1))) == 0
                node2_is_leaf = len(list(G2.graph.successors(node2))) == 0
                #if not ((node1_is_leaf and node1_is_leaf) or (not node1_is_leaf and not node2_is_leaf)): #if node1 is a leaf
                #if (node1_is_leaf and (not node2_is_leaf)) or ( (not node1_is_leaf) and node2_is_leaf):
                if not node1_is_leaf:
                    #only continue if node1 and node2 are both leaves or both not leaves
                    continue

            temp_graph_1 = self.graph.copy()
            temp_graph_1.remove_node(node1)
            remove_nodes_disconnected_from_node(temp_graph_1, self.root)

            #isolating the branch
            branch2 = G2.graph.copy()
            n2_descendants = nx.descendants(branch2,node2)
            for n in list(branch2.nodes):
                if n not in n2_descendants and n is not node2: #removes all nodes not in the branch
                    branch2.remove_node(n)

            branch2 = copy.deepcopy(branch2)
            branch2_root = get_roots(branch2)[0]
            temp_graph_1.add_edges_from(branch2.edges)
            for p in list(self.graph.predecessors(node1)):
                temp_graph_1.add_edge(p,branch2_root)

            if temp_graph_1.number_of_nodes() > self.max_size:
                continue

            self.graph = temp_graph_1

            return True
        return False


    def _crossover_take_branch(self, G2, rng=None):
        '''
        Takes a subgraph from Parent2 and add it to a randomly chosen node in Parent1.
        '''
        rng = np.random.default_rng(rng)

        if self.crossover_same_depth:
            pair_gen = select_nodes_same_depth(self.graph, self.root, G2.graph, G2.root, rng=rng)
        else:
            pair_gen = select_nodes_randomly(self.graph, G2.graph, rng=rng)

        for node1, node2 in pair_gen:
            #TODO: if root is in inner_search_space, then do use it?
            if node2 is G2.root: #dont want to add root as inner node
                continue


            #check if node1 is a leaf and leafs are protected, don't add an input to the leave
            if self.leaf_search_space is not None and len(list(self.graph.successors(node1))) == 0:
                continue

            #icheck if node2 is graph individual
            # if isinstance(node2,GraphIndividual):
            #     if not ((isinstance(node2,GraphIndividual) and ("Recursive" in self.inner_search_space or "Recursive" in self.leaf_search_space))):
            #         continue

            #isolating the branch
            branch2 = G2.graph.copy()
            n2_descendants = nx.descendants(branch2,node2)
            for n in list(branch2.nodes):
                if n not in n2_descendants and n is not node2: #removes all nodes not in the branch
                    branch2.remove_node(n)

            #if node1 plus node2 branch has more than max_children, skip
            if branch2.number_of_nodes() + self.graph.number_of_nodes() > self.max_size:
                continue

            branch2 = copy.deepcopy(branch2)
            branch2_root = get_roots(branch2)[0]
            self.graph.add_edges_from(branch2.edges)
            self.graph.add_edge(node1,branch2_root)

            return True
        return False



    def _crossover_nodes(self, G2, rng=None):
        '''
        Swaps the hyperparamters of one randomly chosen node in Parent1 with the hyperparameters of randomly chosen node in Parent2.
        '''
        rng = np.random.default_rng(rng)

        if self.crossover_same_depth:
            pair_gen = select_nodes_same_depth(self.graph, self.root, G2.graph, G2.root, rng=rng)
        else:
            pair_gen = select_nodes_randomly(self.graph, G2.graph, rng=rng)

        for node1, node2 in pair_gen:

            #if both nodes are leaves
            if len(list(self.graph.successors(node1)))==0 and len(list(G2.graph.successors(node2)))==0:
                if node1.crossover(node2):
                    return True


            #if both nodes are inner nodes
            if len(list(self.graph.successors(node1)))>0 and len(list(G2.graph.successors(node2)))>0:
                if len(list(self.graph.predecessors(node1)))>0 and len(list(G2.graph.predecessors(node2)))>0:
                    if node1.crossover(node2):
                        return True

            #if both nodes are root nodes
            if node1 is self.root and node2 is G2.root:
                if node1.crossover(node2):
                    return True


        return False

    #not including the nodes, just their children
    #Finds leaves attached to nodes and swaps them
    def _crossover_swap_leaf_at_node(self, G2, rng=None):
        rng = np.random.default_rng(rng)

        if self.crossover_same_depth:
            pair_gen = select_nodes_same_depth(self.graph, self.root, G2.graph, G2.root, rng=rng)
        else:
            pair_gen = select_nodes_randomly(self.graph, G2.graph, rng=rng)

        success = False
        for node1, node2 in pair_gen:
            # if leaves are protected node1 and node2 must both be leaves or both be inner nodes
            if self.leaf_search_space is not None and not (len(list(self.graph.successors(node1)))==0 ^ len(list(G2.graph.successors(node2)))==0):
                continue
            #self_leafs = [c for c in nx.descendants(self.graph,node1) if len(list(self.graph.successors(c)))==0 and c is not node1]
            node_leafs = [c for c in nx.descendants(G2.graph,node2) if len(list(G2.graph.successors(c)))==0 and c is not node2]

            # if len(self_leafs) >0:
            #     for c in self_leafs:
            #         if random.choice([True,False]):
            #             self.graph.remove_node(c)
            #             G2.graph.add_edge(node2, c)
            #             success = True

            if len(node_leafs) >0:
                for c in node_leafs:
                    if rng.choice([True,False]):
                        G2.graph.remove_node(c)
                        self.graph.add_edge(node1, c)
                        success = True

        return success



    #TODO edit so that G2 is not modified
    def _crossover_swap_node(self, G2, rng=None):
        '''
        Swaps randomly chosen node from Parent1 with a randomly chosen node from Parent2.
        '''
        rng = np.random.default_rng(rng)

        if self.crossover_same_depth:
            pair_gen = select_nodes_same_depth(self.graph, self.root, G2.graph, G2.root, rng=rng)
        else:
            pair_gen = select_nodes_randomly(self.graph, G2.graph, rng=rng)

        for node1, node2 in pair_gen:
            if node1 is self.root or node2 is G2.root: #TODO: allow root
                continue

            #if leaves are protected
            if self.leaf_search_space is not None:
                #if one node is a leaf, the other must be a leaf
                if not((len(list(self.graph.successors(node1)))==0) ^ (len(list(G2.graph.successors(node2)))==0)):
                    continue #only continue if both are leaves, or both are not leaves


            n1_s = self.graph.successors(node1)
            n1_p = self.graph.predecessors(node1)

            n2_s = G2.graph.successors(node2)
            n2_p = G2.graph.predecessors(node2)

            self.graph.remove_node(node1)
            G2.graph.remove_node(node2)

            self.graph.add_node(node2)

            self.graph.add_edges_from([ (node2, n) for n in n1_s])
            G2.graph.add_edges_from([ (node1, n) for n in n2_s])

            self.graph.add_edges_from([ (n, node2) for n in n1_p])
            G2.graph.add_edges_from([ (n, node1) for n in n2_p])

            return True

        return False


    def _merge_duplicated_nodes(self):

        graph_changed = False
        merged = False
        while(not merged):
            node_list = list(self.graph.nodes)
            merged = True
            for node, other_node in itertools.product(node_list, node_list):
                if node is other_node:
                    continue

                #If nodes are same class/hyperparameters
                if node.unique_id() == other_node.unique_id():
                    node_children = set(self.graph.successors(node))
                    other_node_children = set(self.graph.successors(other_node))
                    #if nodes have identical children, they can be merged
                    if node_children == other_node_children:
                        for other_node_parent in list(self.graph.predecessors(other_node)):
                            if other_node_parent not in self.graph.predecessors(node):
                                self.graph.add_edge(other_node_parent,node)

                        self.graph.remove_node(other_node)
                        merged=False
                        graph_changed = True
                        break

        return graph_changed


    def export_pipeline(self, memory=None, **kwargs):
        estimator_graph = self.graph.copy()

        #mapping = {node:node.method_class(**node.hyperparameters) for node in estimator_graph}
        label_remapping = {}
        label_to_instance = {}

        for node in estimator_graph:
            this_pipeline_node = node.export_pipeline(memory=memory, **kwargs)
            found_unique_label = False
            i=1
            while not found_unique_label:
                label = "{0}_{1}".format(this_pipeline_node.__class__.__name__, i)
                if label not in label_to_instance:
                    found_unique_label = True
                else:
                    i+=1

            label_remapping[node] = label
            label_to_instance[label] = this_pipeline_node

        estimator_graph = nx.relabel_nodes(estimator_graph, label_remapping)

        for label, instance in label_to_instance.items():
            estimator_graph.nodes[label]["instance"] = instance

        return tpot.GraphPipeline(graph=estimator_graph, memory=memory, use_label_encoder=self.use_label_encoder, method=self.method, cross_val_predict_cv=self.cross_val_predict_cv)


    def plot(self):
        G = self.graph.reverse()
        #TODO clean this up
        try:
            pos = nx.planar_layout(G)  # positions for all nodes
        except:
            pos = nx.shell_layout(G)
        # nodes
        options = {'edgecolors': 'tab:gray', 'node_size': 800, 'alpha': 0.9}
        nodelist = list(G.nodes)
        node_color = [plt.cm.Set1(G.nodes[n]['recursive depth']) for n in G]

        fig, ax = plt.subplots()

        nx.draw(G, pos, nodelist=nodelist, node_color=node_color, ax=ax,  **options)


        '''edgelist = []
        for n in n1.node_set:
            for child in n.children:
                edgelist.append((n,child))'''

        # edges
        #nx.draw_networkx_edges(G, pos, width=3.0, arrows=True)
        '''nx.draw_networkx_edges(
            G,
            pos,
            edgelist=[edgelist],
            width=8,
            alpha=0.5,
            edge_color='tab:red',
        )'''



        # some math labels
        labels = {}
        for i, n in enumerate(G.nodes):
            labels[n] = n.method_class.__name__ + "\n" + str(n.hyperparameters)


        nx.draw_networkx_labels(G, pos, labels,ax=ax, font_size=7, font_color='black')

        plt.tight_layout()
        plt.axis('off')
        plt.show()


    def unique_id(self):
        if self.graphkey is None:
            #copy self.graph
            new_graph = self.graph.copy()
            for n in new_graph.nodes:
                new_graph.nodes[n]['label'] = n.unique_id()

            new_graph = nx.convert_node_labels_to_integers(new_graph)
            self.graphkey = GraphKey(new_graph)

        return self.graphkey

crossover(ind2, rng=None)

self 是第一个个体，ind2 是第二个个体。如果 crossover_same_depth 为 True，将选择相同递归深度的图个体进行交叉。否则，将从整个图及其子图中随机选择图个体进行交叉。

这不会影响没有子图的图。也不会影响不是图个体的节点。交叉

源代码位于 tpot/search_spaces/pipelines/graph.py 
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
389
def crossover(self, ind2, rng=None):
    '''
    self is the first individual, ind2 is the second individual
    If crossover_same_depth, it will select graphindividuals at the same recursive depth.
    Otherwise, it will select graphindividuals randomly from the entire graph and its subgraphs.

    This does not impact graphs without subgraphs. And it does not impacts nodes that are not graphindividuals. Cros
    '''

    rng = np.random.default_rng(rng)

    rng.shuffle(self.crossover_methods_list)

    finished = False

    for crossover_method in self.crossover_methods_list:
        if crossover_method(ind2, rng=rng):
            self._merge_duplicated_nodes()
            finished = True
            break

    if self.__debug:
        try:
            nx.find_cycle(self.graph)
            print('something went wrong with ', crossover_method)
        except:
            pass

    if finished:
        self.graphkey = None

    return finished

GraphSearchPipeline

基类：SearchSpace

源代码位于 tpot/search_spaces/pipelines/graph.py 
724
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
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
class GraphSearchPipeline(SearchSpace):
    def __init__(self, 
        root_search_space: SearchSpace, 
        leaf_search_space: SearchSpace = None, 
        inner_search_space: SearchSpace = None, 
        max_size: int = np.inf,
        crossover_same_depth: bool = False,
        cross_val_predict_cv: Union[int, Callable] = 0, #signature function(estimator, X, y=none)
        method: str = 'auto',
        use_label_encoder: bool = False):

        """
        Defines a search space of pipelines in the shape of a Directed Acyclic Graphs. The search spaces for root, leaf, and inner nodes can be defined separately if desired.
        Each graph will have a single root serving as the final estimator which is drawn from the `root_search_space`. If the `leaf_search_space` is defined, all leaves 
        in the pipeline will be drawn from that search space. If the `leaf_search_space` is not defined, all leaves will be drawn from the `inner_search_space`.
        Nodes that are not leaves or roots will be drawn from the `inner_search_space`. If the `inner_search_space` is not defined, there will be no inner nodes.

        `cross_val_predict_cv`, `method`, `memory`, and `use_label_encoder` are passed to the GraphPipeline object when the pipeline is exported and not directly used in the search space.

        Exports to a GraphPipeline object.

        Parameters
        ----------

        root_search_space: SearchSpace
            The search space for the root node of the graph. This node will be the final estimator in the pipeline.

        inner_search_space: SearchSpace, optional
            The search space for the inner nodes of the graph. If not defined, there will be no inner nodes.

        leaf_search_space: SearchSpace, optional
            The search space for the leaf nodes of the graph. If not defined, the leaf nodes will be drawn from the inner_search_space.

        crossover_same_depth: bool, optional
            If True, crossover will only occur between nodes at the same depth in the graph. If False, crossover will occur between nodes at any depth.

        cross_val_predict_cv : int, default=0
            Number of folds to use for the cross_val_predict function for inner classifiers and regressors. Estimators will still be fit on the full dataset, but the following node will get the outputs from cross_val_predict.

            - 0-1 : When set to 0 or 1, the cross_val_predict function will not be used. The next layer will get the outputs from fitting and transforming the full dataset.
            - >=2 : When fitting pipelines with inner classifiers or regressors, they will still be fit on the full dataset.
                    However, the output to the next node will come from cross_val_predict with the specified number of folds.

        method: str, optional
            The prediction method to use for the inner classifiers or regressors. If 'auto', it will try to use predict_proba, decision_function, or predict in that order.

        memory: str or object with the joblib.Memory interface, optional
            Used to cache the input and outputs of nodes to prevent refitting or computationally heavy transformations. By default, no caching is performed. If a string is given, it is the path to the caching directory.

        use_label_encoder: bool, optional
            If True, the label encoder is used to encode the labels to be 0 to N. If False, the label encoder is not used.
            Mainly useful for classifiers (XGBoost) that require labels to be ints from 0 to N.
            Can also be a sklearn.preprocessing.LabelEncoder object. If so, that label encoder is used.

        """


        self.root_search_space = root_search_space
        self.leaf_search_space = leaf_search_space
        self.inner_search_space = inner_search_space
        self.max_size = max_size
        self.crossover_same_depth = crossover_same_depth

        self.cross_val_predict_cv = cross_val_predict_cv
        self.method = method
        self.use_label_encoder = use_label_encoder

    def generate(self, rng=None):
        rng = np.random.default_rng(rng)
        ind =  GraphPipelineIndividual(self.root_search_space, self.leaf_search_space, self.inner_search_space, self.max_size, self.crossover_same_depth, 
                                       self.cross_val_predict_cv, self.method, self.use_label_encoder, rng=rng)  
            # if user specified limit, grab a random number between that limit

        if self.max_size is None or self.max_size == np.inf:
            n_nodes = rng.integers(1, 5)
        else:
            n_nodes = min(rng.integers(1, self.max_size), 5)

        starting_ops = []
        if self.inner_search_space is not None:
            starting_ops.append(ind._mutate_insert_inner_node)
        if self.leaf_search_space is not None or self.inner_search_space is not None:
            starting_ops.append(ind._mutate_insert_leaf)
            n_nodes -= 1

        if len(starting_ops) > 0:
            for _ in range(n_nodes-1):
                func = rng.choice(starting_ops)
                func(rng=rng)

        ind._merge_duplicated_nodes()

        return ind

__init__(root_search_space, leaf_search_space=None, inner_search_space=None, max_size=np.inf, crossover_same_depth=False, cross_val_predict_cv=0, method='auto', use_label_encoder=False)

定义了有向无环图（DAG）形状的管道搜索空间。如果需要，可以单独定义根节点、叶节点和内部节点的搜索空间。每个图将有一个单一根节点作为最终评估器，该节点从 root_search_space 中抽取。如果定义了 leaf_search_space，管道中的所有叶节点将从该搜索空间中抽取。如果未定义 leaf_search_space，所有叶节点将从 inner_search_space 中抽取。非叶节点或根节点的节点将从 inner_search_space 中抽取。如果未定义 inner_search_space，将没有内部节点。

cross_val_predict_cv、method、memory 和 use_label_encoder 在导出管道时传递给 GraphPipeline 对象，并且不直接在搜索空间中使用。

导出为 GraphPipeline 对象。

参数

名称 类型 描述 默认值
root_search_space SearchSpace

图的根节点的搜索空间。此节点将是管道中的最终评估器。

 必需的
inner_search_space SearchSpace

图的内部节点的搜索空间。如果未定义，将没有内部节点。

 None
leaf_search_space SearchSpace

图的叶节点的搜索空间。如果未定义，叶节点将从 inner_search_space 中抽取。

 None
crossover_same_depth bool

如果为 True，交叉操作将仅发生在图中同一深度的节点之间。如果为 False，交叉操作将发生在任何深度的节点之间。

 False
cross_val_predict_cv int

用于内部分类器和回归器的 cross_val_predict 函数的折叠数。评估器仍将在完整数据集上拟合，但下一个节点将从 cross_val_predict 获取输出。

  • 0-1：当设置为 0 或 1 时，不使用 cross_val_predict 函数。下一层将获取在完整数据集上拟合和变换后的输出。

  • ≥2：在拟合包含内部分类器或回归器的管道时，它们仍将在完整数据集上拟合。但是，传递给下一个节点的输出将来自指定折叠数的 cross_val_predict。

 0
method str

用于内部分类器或回归器的预测方法。如果为 'auto'，它将按 predict_proba、decision_function 或 predict 的顺序尝试使用。

 'auto'
memory

用于缓存节点的输入和输出，以防止重复拟合或计算量大的变换。默认情况下，不执行缓存。如果给定字符串，则为缓存目录的路径。

 必需的
use_label_encoder bool

如果为 True，使用标签编码器将标签编码为从 0 到 N。如果为 False，不使用标签编码器。主要用于需要标签为从 0 到 N 的整数的分类器（如 XGBoost）。也可以是 sklearn.preprocessing.LabelEncoder 对象。如果是，则使用该标签编码器。

 False
源代码位于 tpot/search_spaces/pipelines/graph.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
def __init__(self, 
    root_search_space: SearchSpace, 
    leaf_search_space: SearchSpace = None, 
    inner_search_space: SearchSpace = None, 
    max_size: int = np.inf,
    crossover_same_depth: bool = False,
    cross_val_predict_cv: Union[int, Callable] = 0, #signature function(estimator, X, y=none)
    method: str = 'auto',
    use_label_encoder: bool = False):

    """
    Defines a search space of pipelines in the shape of a Directed Acyclic Graphs. The search spaces for root, leaf, and inner nodes can be defined separately if desired.
    Each graph will have a single root serving as the final estimator which is drawn from the `root_search_space`. If the `leaf_search_space` is defined, all leaves 
    in the pipeline will be drawn from that search space. If the `leaf_search_space` is not defined, all leaves will be drawn from the `inner_search_space`.
    Nodes that are not leaves or roots will be drawn from the `inner_search_space`. If the `inner_search_space` is not defined, there will be no inner nodes.

    `cross_val_predict_cv`, `method`, `memory`, and `use_label_encoder` are passed to the GraphPipeline object when the pipeline is exported and not directly used in the search space.

    Exports to a GraphPipeline object.

    Parameters
    ----------

    root_search_space: SearchSpace
        The search space for the root node of the graph. This node will be the final estimator in the pipeline.

    inner_search_space: SearchSpace, optional
        The search space for the inner nodes of the graph. If not defined, there will be no inner nodes.

    leaf_search_space: SearchSpace, optional
        The search space for the leaf nodes of the graph. If not defined, the leaf nodes will be drawn from the inner_search_space.

    crossover_same_depth: bool, optional
        If True, crossover will only occur between nodes at the same depth in the graph. If False, crossover will occur between nodes at any depth.

    cross_val_predict_cv : int, default=0
        Number of folds to use for the cross_val_predict function for inner classifiers and regressors. Estimators will still be fit on the full dataset, but the following node will get the outputs from cross_val_predict.

        - 0-1 : When set to 0 or 1, the cross_val_predict function will not be used. The next layer will get the outputs from fitting and transforming the full dataset.
        - >=2 : When fitting pipelines with inner classifiers or regressors, they will still be fit on the full dataset.
                However, the output to the next node will come from cross_val_predict with the specified number of folds.

    method: str, optional
        The prediction method to use for the inner classifiers or regressors. If 'auto', it will try to use predict_proba, decision_function, or predict in that order.

    memory: str or object with the joblib.Memory interface, optional
        Used to cache the input and outputs of nodes to prevent refitting or computationally heavy transformations. By default, no caching is performed. If a string is given, it is the path to the caching directory.

    use_label_encoder: bool, optional
        If True, the label encoder is used to encode the labels to be 0 to N. If False, the label encoder is not used.
        Mainly useful for classifiers (XGBoost) that require labels to be ints from 0 to N.
        Can also be a sklearn.preprocessing.LabelEncoder object. If so, that label encoder is used.

    """


    self.root_search_space = root_search_space
    self.leaf_search_space = leaf_search_space
    self.inner_search_space = inner_search_space
    self.max_size = max_size
    self.crossover_same_depth = crossover_same_depth

    self.cross_val_predict_cv = cross_val_predict_cv
    self.method = method
    self.use_label_encoder = use_label_encoder

MaskSelector

基类：BaseEstimator, SelectorMixin

选择预定义的特征子集。

源代码位于 tpot/search_spaces/nodes/genetic_feature_selection.py 
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
class MaskSelector(BaseEstimator, SelectorMixin):
    """Select predefined feature subsets."""

    def __init__(self, mask, set_output_transform=None):
        self.mask = mask
        self.set_output_transform = set_output_transform
        if set_output_transform is not None:
            self.set_output(transform=set_output_transform)

    def fit(self, X, y=None):
        self.n_features_in_ = X.shape[1]
        if isinstance(X, pd.DataFrame):
            self.feature_names_in_ = X.columns
        #     self.set_output(transform="pandas")
        self.is_fitted_ = True #so sklearn knows it's fitted
        return self

    def _get_tags(self):
        tags = {"allow_nan": True, "requires_y": False}
        return tags

    def _get_support_mask(self):
        return np.array(self.mask)

    def get_feature_names_out(self, input_features=None):
        return self.feature_names_in_[self.get_support()]

SequentialPipeline

基类：SearchSpace

源代码位于 tpot/search_spaces/pipelines/sequential.py 
174
175
176
177
178
179
180
181
182
183
184
class SequentialPipeline(SearchSpace):
    def __init__(self, search_spaces : List[SearchSpace] ) -> None:
        """
        Takes in a list of search spaces. will produce a pipeline of Sequential length. Each step in the pipeline will correspond to the the search space provided in the same index.
        """

        self.search_spaces = search_spaces

    def generate(self, rng=None):
        rng = np.random.default_rng(rng)
        return SequentialPipelineIndividual(self.search_spaces, rng=rng)

__init__(search_spaces)

接受一个搜索空间列表。将生成一个长度与列表相同的顺序管道。管道中的每个步骤将对应于同一索引中提供的搜索空间。

源代码位于 tpot/search_spaces/pipelines/sequential.py 
175
176
177
178
179
180
def __init__(self, search_spaces : List[SearchSpace] ) -> None:
    """
    Takes in a list of search spaces. will produce a pipeline of Sequential length. Each step in the pipeline will correspond to the the search space provided in the same index.
    """

    self.search_spaces = search_spaces

SklearnIndividual

基类：BaseIndividual

源代码位于 tpot/search_spaces/base.py 
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
class SklearnIndividual(tpot.BaseIndividual):

    def __init_subclass__(cls):
        cls.crossover = cls.validate_same_type(cls.crossover)


    def __init__(self,) -> None:
        super().__init__()

    def mutate(self, rng=None):
        return

    def crossover(self, other, rng=None, **kwargs):
        return 

    @final
    def validate_same_type(func):

        def wrapper(self, other, rng=None, **kwargs):
            if not isinstance(other, type(self)):
                return False
            return func(self, other, rng=rng, **kwargs)

        return wrapper

    def export_pipeline(self, **kwargs) -> BaseEstimator:
        return

    def unique_id(self):
        """
        Returns a unique identifier for the individual. Used for preventing duplicate individuals from being evaluated.
        """
        return self

    #TODO currently TPOT population class manually uses the unique_id to generate the index for the population data frame.
    #alternatively, the index could be the individual itself, with the __eq__ and __hash__ methods implemented.

    # Though this breaks the graphpipeline. When a mutation is called, it changes the __eq__ and __hash__ outputs.
    # Since networkx uses the hash and eq to determine if a node is already in the graph, this causes the graph thing that 
    # This is a new node not in the graph. But this could be changed if when the graphpipeline mutates nodes, 
    # it "replaces" the existing node with the mutated node. This would require a change in the graphpipeline class.

    # def __eq__(self, other):
    #     return self.unique_id() == other.unique_id()

    # def __hash__(self):
    #     return hash(self.unique_id())

    #number of components in the pipeline
    def get_size(self):
        return 1

    @final
    def export_flattened_graphpipeline(self, **graphpipeline_kwargs) -> tpot.GraphPipeline:
        return flatten_to_graphpipeline(self.export_pipeline(), **graphpipeline_kwargs)

unique_id()

返回个体的唯一标识符。用于防止重复个体被评估。

源代码位于 tpot/search_spaces/base.py 
74
75
76
77
78
def unique_id(self):
    """
    Returns a unique identifier for the individual. Used for preventing duplicate individuals from being evaluated.
    """
    return self

TreePipeline

基类：SearchSpace

源代码位于 tpot/search_spaces/pipelines/tree.py 
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
class TreePipeline(SearchSpace):
    def __init__(self, root_search_space : SearchSpace, 
                        leaf_search_space : SearchSpace = None, 
                        inner_search_space : SearchSpace =None, 
                        min_size: int = 2, 
                        max_size: int = 10,
                        crossover_same_depth=False) -> None:

        """
        Generates a pipeline of variable length. Pipeline will have a tree structure similar to TPOT1.

        """

        self.search_space = root_search_space
        self.leaf_search_space = leaf_search_space
        self.inner_search_space = inner_search_space
        self.min_size = min_size
        self.max_size = max_size
        self.crossover_same_depth = crossover_same_depth

    def generate(self, rng=None):
        rng = np.random.default_rng(rng)
        return TreePipelineIndividual(self.search_space, self.leaf_search_space, self.inner_search_space, self.min_size, self.max_size, self.crossover_same_depth, rng=rng)

__init__(root_search_space, leaf_search_space=None, inner_search_space=None, min_size=2, max_size=10, crossover_same_depth=False)

生成变长管道。管道将具有类似于 TPOT1 的树形结构。

源代码位于 tpot/search_spaces/pipelines/tree.py 
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
def __init__(self, root_search_space : SearchSpace, 
                    leaf_search_space : SearchSpace = None, 
                    inner_search_space : SearchSpace =None, 
                    min_size: int = 2, 
                    max_size: int = 10,
                    crossover_same_depth=False) -> None:

    """
    Generates a pipeline of variable length. Pipeline will have a tree structure similar to TPOT1.

    """

    self.search_space = root_search_space
    self.leaf_search_space = leaf_search_space
    self.inner_search_space = inner_search_space
    self.min_size = min_size
    self.max_size = max_size
    self.crossover_same_depth = crossover_same_depth

TupleIndex

TPOT 使用元组为某些管道搜索空间创建唯一 ID。然而，元组有时与 pandas 索引交互不正确。此类是元组的包装器，允许将其用作字典中的键，而不会被视为可迭代对象。

一种替代方案是使唯一 ID 返回字符串，但这不适用于需要特殊对象的 graphpipelines。此类允许线性管道包含图管道，同时仍可作为字典中的键使用。

源代码位于 tpot/search_spaces/tuple_index.py 
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
class TupleIndex():
    """
    TPOT uses tuples to create a unique id for some pipeline search spaces. However, tuples sometimes don't interact correctly with pandas indexes.
    This class is a wrapper around a tuple that allows it to be used as a key in a dictionary, without it being an itereable.

    An alternative could be to make unique id return a string, but this would not work with graphpipelines, which require a special object.
    This class allows linear pipelines to contain graph pipelines while still being able to be used as a key in a dictionary.

    """
    def __init__(self, tup):
        self.tup = tup

    def __eq__(self,other) -> bool:
        return self.tup == other

    def __hash__(self) -> int:
        return self.tup.__hash__()

    def __str__(self) -> str:
        return self.tup.__str__()

    def __repr__(self) -> str:
        return self.tup.__repr__()

UnionPipeline

基类：SearchSpace

源代码位于 tpot/search_spaces/pipelines/union.py 
107
108
109
110
111
112
113
114
115
116
117
class UnionPipeline(SearchSpace):
    def __init__(self, search_spaces : List[SearchSpace] ) -> None:
        """
        Takes in a list of search spaces. will produce a pipeline of Sequential length. Each step in the pipeline will correspond to the the search space provided in the same index.
        """

        self.search_spaces = search_spaces

    def generate(self, rng=None):
        rng = np.random.default_rng(rng)
        return UnionPipelineIndividual(self.search_spaces, rng=rng)

__init__(search_spaces)

接受一个搜索空间列表。将生成一个长度与列表相同的顺序管道。管道中的每个步骤将对应于同一索引中提供的搜索空间。

源代码位于 tpot/search_spaces/pipelines/union.py 
108
109
110
111
112
113
def __init__(self, search_spaces : List[SearchSpace] ) -> None:
    """
    Takes in a list of search spaces. will produce a pipeline of Sequential length. Each step in the pipeline will correspond to the the search space provided in the same index.
    """

    self.search_spaces = search_spaces

UnionPipelineIndividual

基类：SklearnIndividual

接受一个搜索空间列表。列表中的每个项对应于管道中的一个步骤。将生成一个 FeatureUnion 管道。管道中的每个步骤将对应于同一索引中提供的搜索空间。生成的管道将是管道中各步骤的 FeatureUnion。

源代码位于 tpot/search_spaces/pipelines/union.py 
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
class UnionPipelineIndividual(SklearnIndividual):
    """
    Takes in a list of search spaces. each space is a list of SearchSpaces.
    Will produce a FeatureUnion pipeline. Each step in the pipeline will correspond to the the search space provided in the same index.
    The resulting pipeline will be a FeatureUnion of the steps in the pipeline.

    """

    def __init__(self, search_spaces : List[SearchSpace], rng=None) -> None:
        super().__init__()
        self.search_spaces = search_spaces

        self.pipeline = []
        for space in self.search_spaces:
            self.pipeline.append(space.generate(rng))

    def mutate(self, rng=None):
        rng = np.random.default_rng(rng)
        step = rng.choice(self.pipeline)
        return step.mutate(rng)


    def crossover(self, other, rng=None):
        #swap a random step in the pipeline with the corresponding step in the other pipeline
        rng = np.random.default_rng(rng)

        cx_funcs = [self._crossover_node, self._crossover_swap_node]
        rng.shuffle(cx_funcs)
        for cx_func in cx_funcs:
            if cx_func(other, rng):
                return True

        return False

    def _crossover_swap_node(self, other, rng):
        rng = np.random.default_rng(rng)
        idx = rng.integers(1,len(self.pipeline))

        self.pipeline[idx], other.pipeline[idx] = other.pipeline[idx], self.pipeline[idx]
        return True

    def _crossover_node(self, other, rng):
        rng = np.random.default_rng(rng)

        crossover_success = False
        for idx in range(len(self.pipeline)):
            if rng.random() < 0.5:
                if self.pipeline[idx].crossover(other.pipeline[idx], rng):
                    crossover_success = True

        return crossover_success

    def export_pipeline(self, **kwargs):
        return sklearn.pipeline.make_union(*[step.export_pipeline(**kwargs) for step in self.pipeline])

    def unique_id(self):
        l = [step.unique_id() for step in self.pipeline]
        l = ["FeatureUnion"] + l
        return TupleIndex(tuple(l))

WrapperPipeline

基类：SearchSpace

源代码位于 tpot/search_spaces/pipelines/wrapper.py 
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
class WrapperPipeline(SearchSpace):
    def __init__(
            self, 
            method: type, 
            space: ConfigurationSpace,
            estimator_search_space: SearchSpace,
            hyperparameter_parser: callable = None, 
            wrapped_param_name: str = None
            ) -> None:

        """
        This search space is for wrapping a sklearn estimator with a method that takes another estimator and hyperparameters as arguments.
        For example, this can be used with sklearn.ensemble.BaggingClassifier or sklearn.ensemble.AdaBoostClassifier.

        """


        self.estimator_search_space = estimator_search_space
        self.method = method
        self.space = space
        self.hyperparameter_parser=hyperparameter_parser
        self.wrapped_param_name = wrapped_param_name

    def generate(self, rng=None):
        rng = np.random.default_rng(rng)
        return WrapperPipelineIndividual(method=self.method, space=self.space, estimator_search_space=self.estimator_search_space, hyperparameter_parser=self.hyperparameter_parser, wrapped_param_name=self.wrapped_param_name,  rng=rng)

__init__(method, space, estimator_search_space, hyperparameter_parser=None, wrapped_param_name=None)

此搜索空间用于将 sklearn 评估器包装在一个方法中，该方法将另一个评估器和超参数作为参数。例如，可与 sklearn.ensemble.BaggingClassifier 或 sklearn.ensemble.AdaBoostClassifier 一起使用。

源代码位于 tpot/search_spaces/pipelines/wrapper.py 
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
def __init__(
        self, 
        method: type, 
        space: ConfigurationSpace,
        estimator_search_space: SearchSpace,
        hyperparameter_parser: callable = None, 
        wrapped_param_name: str = None
        ) -> None:

    """
    This search space is for wrapping a sklearn estimator with a method that takes another estimator and hyperparameters as arguments.
    For example, this can be used with sklearn.ensemble.BaggingClassifier or sklearn.ensemble.AdaBoostClassifier.

    """


    self.estimator_search_space = estimator_search_space
    self.method = method
    self.space = space
    self.hyperparameter_parser=hyperparameter_parser
    self.wrapped_param_name = wrapped_param_name

get_template_search_spaces(search_space, classification=True, inner_predictors=None, cross_val_predict_cv=None, **get_search_space_params)

返回可由 TPOT 优化的搜索空间。

参数

名称 类型 描述 默认值
search_space

要使用的默认搜索空间。如果为字符串，则应为以下之一：- 'linear'：线性管道的搜索空间 - 'linear-light'：具有更小、更快搜索空间的线性管道搜索空间 - 'graph'：图管道的搜索空间 - 'graph-light'：具有更小、更快搜索空间的图管道搜索空间 - 'mdr'：MDR 管道的搜索空间 如果为 SearchSpace 对象，则应为 TPOT 的有效搜索空间对象。

 必需的
classification

问题是分类问题还是回归问题。

 True
inner_predictors

是否在最终分类器/回归器之前包含额外的分类器/回归器（允许集成）。对于 'linear-light' 和 'graph-light' 搜索空间，默认为 False，否则为 True。（不用于 'mdr' 搜索空间）

 None
cross_val_predict_cv

用于 cross_val_predict 的折叠数。对于 'linear-light' 和 'graph-light' 搜索空间，默认为 0，否则为 5。（不用于 'mdr' 搜索空间）

 None
get_search_space_params

传递给 get_search_space 函数的附加参数。

 {}
源代码位于 tpot/config/template_search_spaces.py 
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
276
277
278
279
280
281
282
283
284
285
286
287
288
def get_template_search_spaces(search_space, classification=True, inner_predictors=None, cross_val_predict_cv=None, **get_search_space_params):
    """
    Returns a search space which can be optimized by TPOT.

    Parameters
    ----------
    search_space: str or SearchSpace
        The default search space to use. If a string, it should be one of the following:
            - 'linear': A search space for linear pipelines
            - 'linear-light': A search space for linear pipelines with a smaller, faster search space
            - 'graph': A search space for graph pipelines
            - 'graph-light': A search space for graph pipelines with a smaller, faster search space
            - 'mdr': A search space for MDR pipelines
        If a SearchSpace object, it should be a valid search space object for TPOT.

    classification: bool, default=True
        Whether the problem is a classification problem or a regression problem.

    inner_predictors: bool, default=None
        Whether to include additional classifiers/regressors before the final classifier/regressor (allowing for ensembles). 
        Defaults to False for 'linear-light' and 'graph-light' search spaces, and True otherwise. (Not used for 'mdr' search space)

    cross_val_predict_cv: int, default=None
        The number of folds to use for cross_val_predict. 
        Defaults to 0 for 'linear-light' and 'graph-light' search spaces, and 5 otherwise. (Not used for 'mdr' search space)

    get_search_space_params: dict
        Additional parameters to pass to the get_search_space function.

    """
    if inner_predictors is None:
        if search_space == "light" or search_space == "graph_light":
            inner_predictors = False
        else:
            inner_predictors = True

    if cross_val_predict_cv is None:
        if search_space == "light" or search_space == "graph_light":
            cross_val_predict_cv = 0
        else:
            if classification:
                cross_val_predict_cv = sklearn.model_selection.StratifiedKFold(n_splits=5, shuffle=True, random_state=42)
            else:
                cross_val_predict_cv = sklearn.model_selection.KFold(n_splits=5, shuffle=True, random_state=42)

    if isinstance(search_space, str):
        if search_space == "linear":
            return get_linear_search_space(classification, inner_predictors, cross_val_predict_cv=cross_val_predict_cv, **get_search_space_params)
        elif search_space == "graph":
            return get_graph_search_space(classification, inner_predictors, cross_val_predict_cv=cross_val_predict_cv, **get_search_space_params)
        elif search_space == "graph-light":
            return get_graph_search_space_light(classification, inner_predictors, cross_val_predict_cv=cross_val_predict_cv, **get_search_space_params)
        elif search_space == "linear-light":
            return get_light_search_space(classification, inner_predictors, cross_val_predict_cv=cross_val_predict_cv, **get_search_space_params)
        elif search_space == "mdr":
            return get_mdr_search_space(classification, **get_search_space_params)
        else:
            raise ValueError("Invalid search space")
    else:
        return search_space