Skip to content

stemflow.model.Hurdle

Hurdle

Bases: BaseEstimator

A simple Hurdle model class

Source code in stemflow/model/Hurdle.py 
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
class Hurdle(BaseEstimator):
    """A simple Hurdle model class"""

    def __init__(self, classifier: BaseEstimator, regressor: BaseEstimator):
        """Make a Hurdle class object

        Args:
            classifier:
                A sklearn style classifier estimator. Must have `fit` and `predict` methods.
                Will be better if it has `predict_proba` method, which returns a numpy array of shape (n_sample, 2)
            regressor:
                A sklearn style regressor estimator. Must have `fit` and `predict` methods.

        Example:
            ```
            >> from xgboost import XGBClassifier, XGBRegressor
            >> from stemflow.model.Hurdle import Hurdle
            >> model = Hurdle(classifier = XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
                              regressor = XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1))
            >> model.fit(X_train, y_train)
            >> pred = model.predict(X_test)
            >> ...

            ```

        """
        self.classifier = classifier
        self.regressor = regressor

    def fit(self, X_train: Union[pd.core.frame.DataFrame, np.ndarray], y_train: Sequence, sample_weight=None):
        """Fitting model

        Args:
            X_train:
                Training variables
            y_train:
                Training target

        """
        binary_ = np.unique(np.where(y_train > 0, 1, 0))
        if len(binary_) == 1:
            # warnings.warn("Warning: only one class presented. Replace with dummy classifier & regressor.")
            self.classifier = dummy_model1(binary_[0])
            self.regressor = dummy_model1(binary_[0])
            return

        new_dat = np.concatenate([np.array(X_train), np.array(y_train).reshape(-1, 1)], axis=1)
        if not isinstance(sample_weight, type(None)):
            self.classifier.fit(new_dat[:, :-1], np.where(new_dat[:, -1] > 0, 1, 0), sample_weight=sample_weight)
        else:
            self.classifier.fit(new_dat[:, :-1], np.where(new_dat[:, -1] > 0, 1, 0))

        regressor_y = new_dat[new_dat[:, -1] > 0, :][:, -1].reshape(-1, 1)
        if regressor_y.shape[0] <= 1:
            self.regressor = dummy_model1(regressor_y[0][0])
        else:
            self.regressor.fit(new_dat[new_dat[:, -1] > 0, :][:, :-1], np.array(regressor_y))

        try:
            self.feature_importances_ = (
                np.array(self.classifier.feature_importances_) + np.array(self.regressor.feature_importances_)
            ) / 2
        except Exception as e:
            warnings.warn(f"Cannot calculate feature importance: {e}")
            pass

    def predict(self, X_test: Union[pd.core.frame.DataFrame, np.ndarray]) -> np.ndarray:
        """Predicting

        Args:
            X_test: Test variables

        Returns:
            A prediction array with shape (-1,1)
        """
        cls_res = self.classifier.predict(X_test)
        reg_res = self.regressor.predict(X_test)
        # reg_res = np.where(reg_res>=0, reg_res, 0) ### we constrain the reg value to be positive
        res = np.where(cls_res > 0, reg_res, cls_res)
        return res.flatten()

    def predict_proba(self, X_test: Union[pd.core.frame.DataFrame, np.ndarray]) -> np.ndarray:
        """Predicting probability

        This method output a numpy array with shape (n_sample, 2)
        However, user should notice that this is only for structuring the sklearn predict_proba-like method
        Only the res[:,1] is meaningful, aka the last dimension in the two dimensions. The first dimension is always zero.

        Args:
            X_test:
                Testing variables

        Returns:
            Prediction results with shape (n_samples, 2)
        """
        a = np.zeros(len(X_test)).reshape(-1, 1)
        b = self.predict(X_test).reshape(-1, 1)
        res = np.concatenate([a, b], axis=1)
        return res

__init__(classifier, regressor)

Make a Hurdle class object

Parameters:

  • classifier (BaseEstimator) –

    A sklearn style classifier estimator. Must have fit and predict methods. Will be better if it has predict_proba method, which returns a numpy array of shape (n_sample, 2)

  • regressor (BaseEstimator) –

    A sklearn style regressor estimator. Must have fit and predict methods.

Example 
>> from xgboost import XGBClassifier, XGBRegressor
>> from stemflow.model.Hurdle import Hurdle
>> model = Hurdle(classifier = XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
                  regressor = XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1))
>> model.fit(X_train, y_train)
>> pred = model.predict(X_test)
>> ...
Source code in stemflow/model/Hurdle.py 
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
def __init__(self, classifier: BaseEstimator, regressor: BaseEstimator):
    """Make a Hurdle class object

    Args:
        classifier:
            A sklearn style classifier estimator. Must have `fit` and `predict` methods.
            Will be better if it has `predict_proba` method, which returns a numpy array of shape (n_sample, 2)
        regressor:
            A sklearn style regressor estimator. Must have `fit` and `predict` methods.

    Example:
        ```
        >> from xgboost import XGBClassifier, XGBRegressor
        >> from stemflow.model.Hurdle import Hurdle
        >> model = Hurdle(classifier = XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
                          regressor = XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1))
        >> model.fit(X_train, y_train)
        >> pred = model.predict(X_test)
        >> ...

        ```

    """
    self.classifier = classifier
    self.regressor = regressor

fit(X_train, y_train, sample_weight=None)

Fitting model

Parameters:

  • X_train (Union[DataFrame, ndarray]) –

    Training variables

  • y_train (Sequence) –

    Training target

Source code in stemflow/model/Hurdle.py 
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
def fit(self, X_train: Union[pd.core.frame.DataFrame, np.ndarray], y_train: Sequence, sample_weight=None):
    """Fitting model

    Args:
        X_train:
            Training variables
        y_train:
            Training target

    """
    binary_ = np.unique(np.where(y_train > 0, 1, 0))
    if len(binary_) == 1:
        # warnings.warn("Warning: only one class presented. Replace with dummy classifier & regressor.")
        self.classifier = dummy_model1(binary_[0])
        self.regressor = dummy_model1(binary_[0])
        return

    new_dat = np.concatenate([np.array(X_train), np.array(y_train).reshape(-1, 1)], axis=1)
    if not isinstance(sample_weight, type(None)):
        self.classifier.fit(new_dat[:, :-1], np.where(new_dat[:, -1] > 0, 1, 0), sample_weight=sample_weight)
    else:
        self.classifier.fit(new_dat[:, :-1], np.where(new_dat[:, -1] > 0, 1, 0))

    regressor_y = new_dat[new_dat[:, -1] > 0, :][:, -1].reshape(-1, 1)
    if regressor_y.shape[0] <= 1:
        self.regressor = dummy_model1(regressor_y[0][0])
    else:
        self.regressor.fit(new_dat[new_dat[:, -1] > 0, :][:, :-1], np.array(regressor_y))

    try:
        self.feature_importances_ = (
            np.array(self.classifier.feature_importances_) + np.array(self.regressor.feature_importances_)
        ) / 2
    except Exception as e:
        warnings.warn(f"Cannot calculate feature importance: {e}")
        pass

predict(X_test)

Predicting

Parameters:

  • X_test (Union[DataFrame, ndarray]) –

    Test variables

Returns:

  • ndarray –

    A prediction array with shape (-1,1)

Source code in stemflow/model/Hurdle.py 
78
79
80
81
82
83
84
85
86
87
88
89
90
91
def predict(self, X_test: Union[pd.core.frame.DataFrame, np.ndarray]) -> np.ndarray:
    """Predicting

    Args:
        X_test: Test variables

    Returns:
        A prediction array with shape (-1,1)
    """
    cls_res = self.classifier.predict(X_test)
    reg_res = self.regressor.predict(X_test)
    # reg_res = np.where(reg_res>=0, reg_res, 0) ### we constrain the reg value to be positive
    res = np.where(cls_res > 0, reg_res, cls_res)
    return res.flatten()

predict_proba(X_test)

Predicting probability

This method output a numpy array with shape (n_sample, 2) However, user should notice that this is only for structuring the sklearn predict_proba-like method Only the res[:,1] is meaningful, aka the last dimension in the two dimensions. The first dimension is always zero.

Parameters:

  • X_test (Union[DataFrame, ndarray]) –

    Testing variables

Returns:

  • ndarray –

    Prediction results with shape (n_samples, 2)

Source code in stemflow/model/Hurdle.py 
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
def predict_proba(self, X_test: Union[pd.core.frame.DataFrame, np.ndarray]) -> np.ndarray:
    """Predicting probability

    This method output a numpy array with shape (n_sample, 2)
    However, user should notice that this is only for structuring the sklearn predict_proba-like method
    Only the res[:,1] is meaningful, aka the last dimension in the two dimensions. The first dimension is always zero.

    Args:
        X_test:
            Testing variables

    Returns:
        Prediction results with shape (n_samples, 2)
    """
    a = np.zeros(len(X_test)).reshape(-1, 1)
    b = self.predict(X_test).reshape(-1, 1)
    res = np.concatenate([a, b], axis=1)
    return res

Hurdle_for_AdaSTEM

Bases: BaseEstimator

Source code in stemflow/model/Hurdle.py 
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
class Hurdle_for_AdaSTEM(BaseEstimator):
    def __init__(self, classifier: BaseEstimator, regressor: BaseEstimator):
        """Make a Hurdle_for_AdaSTEM class object

        Normally speaking, AdaSTEMClassifier and AdaSTEMRegressor should be passed here if using this class.

        Args:
            classifier:
                A sklearn style classifier estimator (should be AdaSTEMClassifier here). Must have `fit` and `predict` methods.
                Will be better if it has `predict_proba` method, which returns a numpy array of shape (n_sample, 2)
            regressor:
                A sklearn style regressor estimator (should be AdaSTEMRegressor here). Must have `fit` and `predict` methods.

        Example:
            ```
            >> from stemflow.model.AdaSTEM import AdaSTEM, AdaSTEMClassifier, AdaSTEMRegressor
            >> from stemflow.model.Hurdle import Hurdle_for_AdaSTEM
            >> from xgboost import XGBClassifier, XGBRegressor

            >> SAVE_DIR = './'

            >> model = Hurdle_for_AdaSTEM(
            ...     classifier=AdaSTEMClassifier(base_model=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
            ...                                 save_gridding_plot = True,
            ...                                 ensemble_fold=10,
            ...                                 min_ensemble_required=7,
            ...                                 grid_len_lon_upper_threshold=25,
            ...                                 grid_len_lon_lower_threshold=5,
            ...                                 grid_len_lat_upper_threshold=25,
            ...                                 grid_len_lat_lower_threshold=5,
            ...                                 points_lower_threshold=50,
            ...                                 Spatio1='longitude',
            ...                                 Spatio2 = 'latitude',
            ...                                 Temporal1 = 'DOY',
            ...                                 use_temporal_to_train=True),
            ...     regressor=AdaSTEMRegressor(base_model=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
            ...                                 save_gridding_plot = True,
            ...                                 ensemble_fold=10,
            ...                                 min_ensemble_required=7,
            ...                                 grid_len_lon_upper_threshold=25,
            ...                                 grid_len_lon_lower_threshold=5,
            ...                                 grid_len_lat_upper_threshold=25,
            ...                                 grid_len_lat_lower_threshold=5,
            ...                                 points_lower_threshold=50,
            ...                                 Spatio1='longitude',
            ...                                 Spatio2 = 'latitude',
            ...                                 Temporal1 = 'DOY',
            ...                                 use_temporal_to_train=True)
            ... )

            >> ## fit
            >> model.fit(X_train.reset_index(drop=True), y_train)

            >> ## predict
            >> pred = model.predict(X_test)
            >> pred = np.where(pred<0, 0, pred)
            >> eval_metrics = AdaSTEM.eval_STEM_res('hurdle',y_test, pred_mean)
            >> print(eval_metrics)


            ```

        """

        self.classifier = classifier
        self.regressor = regressor

    def fit(self, X_train: Union[pd.core.frame.DataFrame, np.ndarray], y_train: Sequence, verbosity: int = 1):
        """Fitting model
        Args:
            X_train:
                Training variables
            y_train:
                Training target
            verbosity:
                Whether to show progress bar. 0 for No, and Yes other wise.

        """
        binary_ = np.unique(np.where(y_train > 0, 1, 0))
        if len(binary_) == 1:
            warnings.warn("Warning: only one class presented. Replace with dummy classifier & regressor.")
            self.classifier = dummy_model1(binary_[0])
            self.regressor = dummy_model1(binary_[0])
            return

        X_train["y_train"] = y_train

        if verbosity == 0:
            self.classifier.fit(X_train.iloc[:, :-1], np.where(X_train.iloc[:, -1].values > 0, 1, 0), verbosity=0)
            self.regressor.fit(
                X_train[X_train["y_train"] > 0].iloc[:, :-1],
                np.array(X_train[X_train["y_train"] > 0].iloc[:, -1]),
                verbosity=0,
            )
        else:
            self.classifier.fit(X_train.iloc[:, :-1], np.where(X_train.iloc[:, -1].values > 0, 1, 0), verbosity=1)
            self.regressor.fit(
                X_train[X_train["y_train"] > 0].iloc[:, :-1],
                np.array(X_train[X_train["y_train"] > 0].iloc[:, -1]),
                verbosity=1,
            )

        return self

    def predict(
        self,
        X_test: Union[pd.core.frame.DataFrame, np.ndarray],
        n_jobs: int = 1,
        verbosity: int = 1,
        return_by_separate_ensembles: bool = False,
    ) -> np.ndarray:
        """Predict

        Args:
            X_test:
                Test variables
            n_jobs:
                Multi-processing in prediction.
            verbosity:
                Whether to show progress bar. 0 for No, and Yes other wise.
            return_by_separate_ensembles (bool, optional):
                Test function. return not by aggregation, but by separate ensembles.

        Returns:
            A prediction array with shape (-1,1)
        """
        if verbosity == 0:
            cls_res = self.classifier.predict(
                X_test, n_jobs=n_jobs, verbosity=0, return_by_separate_ensembles=return_by_separate_ensembles
            )
            reg_res = self.regressor.predict(
                X_test, n_jobs=n_jobs, verbosity=0, return_by_separate_ensembles=return_by_separate_ensembles
            )
        else:
            cls_res = self.classifier.predict(
                X_test, n_jobs=n_jobs, verbosity=1, return_by_separate_ensembles=return_by_separate_ensembles
            )
            reg_res = self.regressor.predict(
                X_test, n_jobs=n_jobs, verbosity=1, return_by_separate_ensembles=return_by_separate_ensembles
            )
        # reg_res = np.where(reg_res>=0, reg_res, 0) ### we constrain the reg value to be positive
        res = np.where(cls_res < 0.5, 0, cls_res)
        res = np.where(cls_res > 0.5, reg_res, cls_res)
        return res.flatten()

    def predict_proba(
        self,
        X_test: Union[pd.core.frame.DataFrame, np.ndarray],
        n_jobs: int = 1,
        verbosity: int = 0,
        return_by_separate_ensembles: bool = False,
    ) -> np.ndarray:
        """Just a rewrite of `predict` method

        Args:
            X_test:
                Testing variables
            n_jobs:
                Multi-processing in prediction.
            verbosity:
                Whether to show progress bar. 0 for No, and Yes other wise.
            return_by_separate_ensembles (bool, optional):
                Test function. return not by aggregation, but by separate ensembles.

        Returns:
            A prediction array with shape (-1,1)
        """

        return self.predict(
            self, X_test, n_jobs=n_jobs, verbosity=verbosity, return_by_separate_ensembles=return_by_separate_ensembles
        )

__init__(classifier, regressor)

Make a Hurdle_for_AdaSTEM class object

Normally speaking, AdaSTEMClassifier and AdaSTEMRegressor should be passed here if using this class.

Parameters:

  • classifier (BaseEstimator) –

    A sklearn style classifier estimator (should be AdaSTEMClassifier here). Must have fit and predict methods. Will be better if it has predict_proba method, which returns a numpy array of shape (n_sample, 2)

  • regressor (BaseEstimator) –

    A sklearn style regressor estimator (should be AdaSTEMRegressor here). Must have fit and predict methods.

Example 
>> from stemflow.model.AdaSTEM import AdaSTEM, AdaSTEMClassifier, AdaSTEMRegressor
>> from stemflow.model.Hurdle import Hurdle_for_AdaSTEM
>> from xgboost import XGBClassifier, XGBRegressor

>> SAVE_DIR = './'

>> model = Hurdle_for_AdaSTEM(
...     classifier=AdaSTEMClassifier(base_model=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
...                                 save_gridding_plot = True,
...                                 ensemble_fold=10,
...                                 min_ensemble_required=7,
...                                 grid_len_lon_upper_threshold=25,
...                                 grid_len_lon_lower_threshold=5,
...                                 grid_len_lat_upper_threshold=25,
...                                 grid_len_lat_lower_threshold=5,
...                                 points_lower_threshold=50,
...                                 Spatio1='longitude',
...                                 Spatio2 = 'latitude',
...                                 Temporal1 = 'DOY',
...                                 use_temporal_to_train=True),
...     regressor=AdaSTEMRegressor(base_model=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
...                                 save_gridding_plot = True,
...                                 ensemble_fold=10,
...                                 min_ensemble_required=7,
...                                 grid_len_lon_upper_threshold=25,
...                                 grid_len_lon_lower_threshold=5,
...                                 grid_len_lat_upper_threshold=25,
...                                 grid_len_lat_lower_threshold=5,
...                                 points_lower_threshold=50,
...                                 Spatio1='longitude',
...                                 Spatio2 = 'latitude',
...                                 Temporal1 = 'DOY',
...                                 use_temporal_to_train=True)
... )

>> ## fit
>> model.fit(X_train.reset_index(drop=True), y_train)

>> ## predict
>> pred = model.predict(X_test)
>> pred = np.where(pred<0, 0, pred)
>> eval_metrics = AdaSTEM.eval_STEM_res('hurdle',y_test, pred_mean)
>> print(eval_metrics)
Source code in stemflow/model/Hurdle.py 
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
def __init__(self, classifier: BaseEstimator, regressor: BaseEstimator):
    """Make a Hurdle_for_AdaSTEM class object

    Normally speaking, AdaSTEMClassifier and AdaSTEMRegressor should be passed here if using this class.

    Args:
        classifier:
            A sklearn style classifier estimator (should be AdaSTEMClassifier here). Must have `fit` and `predict` methods.
            Will be better if it has `predict_proba` method, which returns a numpy array of shape (n_sample, 2)
        regressor:
            A sklearn style regressor estimator (should be AdaSTEMRegressor here). Must have `fit` and `predict` methods.

    Example:
        ```
        >> from stemflow.model.AdaSTEM import AdaSTEM, AdaSTEMClassifier, AdaSTEMRegressor
        >> from stemflow.model.Hurdle import Hurdle_for_AdaSTEM
        >> from xgboost import XGBClassifier, XGBRegressor

        >> SAVE_DIR = './'

        >> model = Hurdle_for_AdaSTEM(
        ...     classifier=AdaSTEMClassifier(base_model=XGBClassifier(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
        ...                                 save_gridding_plot = True,
        ...                                 ensemble_fold=10,
        ...                                 min_ensemble_required=7,
        ...                                 grid_len_lon_upper_threshold=25,
        ...                                 grid_len_lon_lower_threshold=5,
        ...                                 grid_len_lat_upper_threshold=25,
        ...                                 grid_len_lat_lower_threshold=5,
        ...                                 points_lower_threshold=50,
        ...                                 Spatio1='longitude',
        ...                                 Spatio2 = 'latitude',
        ...                                 Temporal1 = 'DOY',
        ...                                 use_temporal_to_train=True),
        ...     regressor=AdaSTEMRegressor(base_model=XGBRegressor(tree_method='hist',random_state=42, verbosity = 0, n_jobs=1),
        ...                                 save_gridding_plot = True,
        ...                                 ensemble_fold=10,
        ...                                 min_ensemble_required=7,
        ...                                 grid_len_lon_upper_threshold=25,
        ...                                 grid_len_lon_lower_threshold=5,
        ...                                 grid_len_lat_upper_threshold=25,
        ...                                 grid_len_lat_lower_threshold=5,
        ...                                 points_lower_threshold=50,
        ...                                 Spatio1='longitude',
        ...                                 Spatio2 = 'latitude',
        ...                                 Temporal1 = 'DOY',
        ...                                 use_temporal_to_train=True)
        ... )

        >> ## fit
        >> model.fit(X_train.reset_index(drop=True), y_train)

        >> ## predict
        >> pred = model.predict(X_test)
        >> pred = np.where(pred<0, 0, pred)
        >> eval_metrics = AdaSTEM.eval_STEM_res('hurdle',y_test, pred_mean)
        >> print(eval_metrics)


        ```

    """

    self.classifier = classifier
    self.regressor = regressor

fit(X_train, y_train, verbosity=1)

Fitting model Args: X_train: Training variables y_train: Training target verbosity: Whether to show progress bar. 0 for No, and Yes other wise.

Source code in stemflow/model/Hurdle.py 
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
def fit(self, X_train: Union[pd.core.frame.DataFrame, np.ndarray], y_train: Sequence, verbosity: int = 1):
    """Fitting model
    Args:
        X_train:
            Training variables
        y_train:
            Training target
        verbosity:
            Whether to show progress bar. 0 for No, and Yes other wise.

    """
    binary_ = np.unique(np.where(y_train > 0, 1, 0))
    if len(binary_) == 1:
        warnings.warn("Warning: only one class presented. Replace with dummy classifier & regressor.")
        self.classifier = dummy_model1(binary_[0])
        self.regressor = dummy_model1(binary_[0])
        return

    X_train["y_train"] = y_train

    if verbosity == 0:
        self.classifier.fit(X_train.iloc[:, :-1], np.where(X_train.iloc[:, -1].values > 0, 1, 0), verbosity=0)
        self.regressor.fit(
            X_train[X_train["y_train"] > 0].iloc[:, :-1],
            np.array(X_train[X_train["y_train"] > 0].iloc[:, -1]),
            verbosity=0,
        )
    else:
        self.classifier.fit(X_train.iloc[:, :-1], np.where(X_train.iloc[:, -1].values > 0, 1, 0), verbosity=1)
        self.regressor.fit(
            X_train[X_train["y_train"] > 0].iloc[:, :-1],
            np.array(X_train[X_train["y_train"] > 0].iloc[:, -1]),
            verbosity=1,
        )

    return self

predict(X_test, n_jobs=1, verbosity=1, return_by_separate_ensembles=False)

Predict

Parameters:

  • X_test (Union[DataFrame, ndarray]) –

    Test variables

  • n_jobs (int, default: 1 ) –

    Multi-processing in prediction.

  • verbosity (int, default: 1 ) –

    Whether to show progress bar. 0 for No, and Yes other wise.

  • return_by_separate_ensembles (bool, default: False ) –

    Test function. return not by aggregation, but by separate ensembles.

Returns:

  • ndarray –

    A prediction array with shape (-1,1)

Source code in stemflow/model/Hurdle.py 
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
def predict(
    self,
    X_test: Union[pd.core.frame.DataFrame, np.ndarray],
    n_jobs: int = 1,
    verbosity: int = 1,
    return_by_separate_ensembles: bool = False,
) -> np.ndarray:
    """Predict

    Args:
        X_test:
            Test variables
        n_jobs:
            Multi-processing in prediction.
        verbosity:
            Whether to show progress bar. 0 for No, and Yes other wise.
        return_by_separate_ensembles (bool, optional):
            Test function. return not by aggregation, but by separate ensembles.

    Returns:
        A prediction array with shape (-1,1)
    """
    if verbosity == 0:
        cls_res = self.classifier.predict(
            X_test, n_jobs=n_jobs, verbosity=0, return_by_separate_ensembles=return_by_separate_ensembles
        )
        reg_res = self.regressor.predict(
            X_test, n_jobs=n_jobs, verbosity=0, return_by_separate_ensembles=return_by_separate_ensembles
        )
    else:
        cls_res = self.classifier.predict(
            X_test, n_jobs=n_jobs, verbosity=1, return_by_separate_ensembles=return_by_separate_ensembles
        )
        reg_res = self.regressor.predict(
            X_test, n_jobs=n_jobs, verbosity=1, return_by_separate_ensembles=return_by_separate_ensembles
        )
    # reg_res = np.where(reg_res>=0, reg_res, 0) ### we constrain the reg value to be positive
    res = np.where(cls_res < 0.5, 0, cls_res)
    res = np.where(cls_res > 0.5, reg_res, cls_res)
    return res.flatten()

predict_proba(X_test, n_jobs=1, verbosity=0, return_by_separate_ensembles=False)

Just a rewrite of predict method

Parameters:

  • X_test (Union[DataFrame, ndarray]) –

    Testing variables

  • n_jobs (int, default: 1 ) –

    Multi-processing in prediction.

  • verbosity (int, default: 0 ) –

    Whether to show progress bar. 0 for No, and Yes other wise.

  • return_by_separate_ensembles (bool, default: False ) –

    Test function. return not by aggregation, but by separate ensembles.

Returns:

  • ndarray –

    A prediction array with shape (-1,1)

Source code in stemflow/model/Hurdle.py 
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
def predict_proba(
    self,
    X_test: Union[pd.core.frame.DataFrame, np.ndarray],
    n_jobs: int = 1,
    verbosity: int = 0,
    return_by_separate_ensembles: bool = False,
) -> np.ndarray:
    """Just a rewrite of `predict` method

    Args:
        X_test:
            Testing variables
        n_jobs:
            Multi-processing in prediction.
        verbosity:
            Whether to show progress bar. 0 for No, and Yes other wise.
        return_by_separate_ensembles (bool, optional):
            Test function. return not by aggregation, but by separate ensembles.

    Returns:
        A prediction array with shape (-1,1)
    """

    return self.predict(
        self, X_test, n_jobs=n_jobs, verbosity=verbosity, return_by_separate_ensembles=return_by_separate_ensembles
    )