随机搜索交叉验证：所有评估器均无法拟合。

Question

我目前正在参加 Kaggle 上的 "French Motor Claims Datasets freMTPL2freq" 技术竞赛（https://www.kaggle.com/floser/french-motor-claims-datasets-fremtpl2freq）。不幸的是，每当我使用 RandomizedSearchCV 时，都会出现"NotFittedError: All estimators failed to fit"错误，我无法弄清楚原因。如有任何帮助，将不胜感激。

import numpy as np
import statsmodels.api as sm
import scipy.stats as stats
from matplotlib import pyplot as plt
from sklearn.pipeline import Pipeline
from sklearn.compose import ColumnTransformer
from sklearn.preprocessing import OneHotEncoder
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import KBinsDiscretizer
from sklearn.model_selection import train_test_split
from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import RandomizedSearchCV
from sklearn.model_selection import StratifiedKFold
from sklearn.metrics import mean_poisson_deviance
from sklearn.metrics import mean_squared_error
from sklearn.ensemble import VotingRegressor
from sklearn.ensemble import StackingRegressor
from sklearn.metrics import mean_gamma_deviance
from sklearn.metrics import mean_squared_error
from xgboost import XGBRegressor

data_freq = pd.read_csv('freMTPL2freq.csv')
data_freq['Area'] = data_freq['Area'].str.replace('\'','')
data_freq['VehBrand'] = data_freq['VehBrand'].str.replace('\'','')
data_freq['VehGas'] = data_freq['VehGas'].str.replace('\'','')
data_freq['Region'] = data_freq['Region'].str.replace('\'','')

    data_freq['frequency'] = data_freq['ClaimNb'] / data_freq['Exposure']
    y = data_freq['frequency']
    X = data_freq.drop(['frequency', 'ClaimNb', 'IDpol'], axis = 1)
    X_train, X_val, y_train, y_val = train_test_split(X,y, test_size=0.2, shuffle = True, random_state = 42)
    pt_columns = ['VehPower', 'VehAge', 'DrivAge', 'BonusMalus', 'Density']
    cat_columns = ['Area', 'Region', 'VehBrand', 'VehGas']

    from xgboost import XGBRegressor
    ct = ColumnTransformer([('pt', 'passthrough', pt_columns),
                       ('ohe', OneHotEncoder(), cat_columns)])
    pipe_xgbr = Pipeline([('cf_trans', ct),
                      ('ssc', StandardScaler(with_mean = False)),
    ('xgb_regressor', XGBRegressor())
    ])

    param = {'xgb_regressor__n_estimators':[3, 5],
        'xgb_regressor__max_depth':[3, 5, 7],
        'xgb_regressor__learning_rate':[0.1, 0.5],
        'xgb_regressor__colsample_bytree':[0.5, 0.8],
        'xgb_regressor__subsample':[0.5, 0.8]
        }

    rscv = RandomizedSearchCV(pipe_xgbr, param_distributions = param, n_iter = 2, scoring = mean_squared_error, n_jobs = -1, cv = 5, error_score = 'raise')
    rscv.fit(X_train, y_train, xgbr_regressor__sample_weight = X_train['Exposure'])

原始数据框 data_freq 的前五行如下：

   IDpol    ClaimNb Exposure    Area    VehPower    VehAge  DrivAge BonusMalus  VehBrand    VehGas  Density Region
0   1.0        1        0.10       D           5         0       55        50        B12    Regular 1217    R82
1   3.0        1        0.77       D           5         0       55        50        B12    Regular 1217    R82
2   5.0        1        0.75       B           6         2       52        50        B12    Diesel  54      R22
3   10.0       1        0.09       B           7         0       46        50        B12    Diesel  76      R72
4   11.0       1        0.84       B           7         0       46        50        B12    Diesel  76      R72

我遇到的错误如下：

---------------------------------------------------------------------------
_RemoteTraceback                          Traceback (most recent call last)
_RemoteTraceback: 
"""
Traceback (most recent call last):
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\externals\loky\process_executor.py", line 418, in _process_worker
    r = call_item()
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\externals\loky\process_executor.py", line 272, in __call__
    return self.fn(*self.args, **self.kwargs)
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\_parallel_backends.py", line 608, in __call__
    return self.func(*args, **kwargs)
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\parallel.py", line 256, in __call__
    for func, args, kwargs in self.items]
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\parallel.py", line 256, in <listcomp>
    for func, args, kwargs in self.items]
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\utils\fixes.py", line 222, in __call__
    return self.function(*args, **kwargs)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\model_selection\_validation.py", line 598, in _fit_and_score
    estimator.fit(X_train, y_train, **fit_params)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\pipeline.py", line 340, in fit
    fit_params_steps = self._check_fit_params(**fit_params)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\pipeline.py", line 261, in _check_fit_params
    fit_params_steps[step][param] = pval
KeyError: 'xgbr_regressor'
"""

The above exception was the direct cause of the following exception:

KeyError                                  Traceback (most recent call last)
<ipython-input-68-0c1886d1e985> in <module>
----> 1 rscv.fit(X_train, y_train, xgbr_regressor__sample_weight = X_train['Exposure'])
      2 #pipe_xgbr.fit(X_train, y_train)
      3 #X_train.describe(include = 'all')

~\anaconda3\lib\site-packages\sklearn\utils\validation.py in inner_f(*args, **kwargs)
     61             extra_args = len(args) - len(all_args)
     62             if extra_args <= 0:
---> 63                 return f(*args, **kwargs)
     64 
     65             # extra_args > 0

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in fit(self, X, y, groups, **fit_params)
    839                 return results
    840 
--> 841             self._run_search(evaluate_candidates)
    842 
    843             # multimetric is determined here because in the case of a callable

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in _run_search(self, evaluate_candidates)
   1633         evaluate_candidates(ParameterSampler(
   1634             self.param_distributions, self.n_iter,
-> 1635             random_state=self.random_state))

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in evaluate_candidates(candidate_params, cv, more_results)
    807                                    (split_idx, (train, test)) in product(
    808                                    enumerate(candidate_params),
--> 809                                    enumerate(cv.split(X, y, groups))))
    810 
    811                 if len(out) < 1:

~\anaconda3\lib\site-packages\joblib\parallel.py in __call__(self, iterable)
   1015 
   1016             with self._backend.retrieval_context():
-> 1017                 self.retrieve()
   1018             # Make sure that we get a last message telling us we are done
   1019             elapsed_time = time.time() - self._start_time

~\anaconda3\lib\site-packages\joblib\parallel.py in retrieve(self)
    907             try:
    908                 if getattr(self._backend, 'supports_timeout', False):
--> 909                     self._output.extend(job.get(timeout=self.timeout))
    910                 else:
    911                     self._output.extend(job.get())

~\anaconda3\lib\site-packages\joblib\_parallel_backends.py in wrap_future_result(future, timeout)
    560         AsyncResults.get from multiprocessing."""
    561         try:
--> 562             return future.result(timeout=timeout)
    563         except LokyTimeoutError:
    564             raise TimeoutError()

~\anaconda3\lib\concurrent\futures\_base.py in result(self, timeout)
    433                 raise CancelledError()
    434             elif self._state == FINISHED:
--> 435                 return self.__get_result()
    436             else:
    437                 raise TimeoutError()

~\anaconda3\lib\concurrent\futures\_base.py in __get_result(self)
    382     def __get_result(self):
    383         if self._exception:
--> 384             raise self._exception
    385         else:
    386             return self._result

KeyError: 'xgbr_regressor'

我还尝试在没有使用样本权重参数的情况下运行fit。 在这种情况下，错误会发生变化：

---------------------------------------------------------------------------
_RemoteTraceback                          Traceback (most recent call last)
_RemoteTraceback: 
"""
Traceback (most recent call last):
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\externals\loky\process_executor.py", line 418, in _process_worker
    r = call_item()
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\externals\loky\process_executor.py", line 272, in __call__
    return self.fn(*self.args, **self.kwargs)
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\_parallel_backends.py", line 608, in __call__
    return self.func(*args, **kwargs)
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\parallel.py", line 256, in __call__
    for func, args, kwargs in self.items]
  File "C:\Users\Jan\anaconda3\lib\site-packages\joblib\parallel.py", line 256, in <listcomp>
    for func, args, kwargs in self.items]
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\utils\fixes.py", line 222, in __call__
    return self.function(*args, **kwargs)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\model_selection\_validation.py", line 625, in _fit_and_score
    test_scores = _score(estimator, X_test, y_test, scorer, error_score)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\model_selection\_validation.py", line 687, in _score
    scores = scorer(estimator, X_test, y_test)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\utils\validation.py", line 74, in inner_f
    return f(**kwargs)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\metrics\_regression.py", line 336, in mean_squared_error
    y_true, y_pred, multioutput)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\metrics\_regression.py", line 88, in _check_reg_targets
    check_consistent_length(y_true, y_pred)
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\utils\validation.py", line 316, in check_consistent_length
    lengths = [_num_samples(X) for X in arrays if X is not None]
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\utils\validation.py", line 316, in <listcomp>
    lengths = [_num_samples(X) for X in arrays if X is not None]
  File "C:\Users\Jan\anaconda3\lib\site-packages\sklearn\utils\validation.py", line 249, in _num_samples
    raise TypeError(message)
TypeError: Expected sequence or array-like, got <class 'sklearn.pipeline.Pipeline'>
"""

The above exception was the direct cause of the following exception:

TypeError                                 Traceback (most recent call last)
<ipython-input-69-a9be9cc5df4a> in <module>
----> 1 rscv.fit(X_train, y_train)#, xgbr_regressor__sample_weight = X_train['Exposure'])
      2 #pipe_xgbr.fit(X_train, y_train)
      3 #X_train.describe(include = 'all')

~\anaconda3\lib\site-packages\sklearn\utils\validation.py in inner_f(*args, **kwargs)
     61             extra_args = len(args) - len(all_args)
     62             if extra_args <= 0:
---> 63                 return f(*args, **kwargs)
     64 
     65             # extra_args > 0

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in fit(self, X, y, groups, **fit_params)
    839                 return results
    840 
--> 841             self._run_search(evaluate_candidates)
    842 
    843             # multimetric is determined here because in the case of a callable

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in _run_search(self, evaluate_candidates)
   1633         evaluate_candidates(ParameterSampler(
   1634             self.param_distributions, self.n_iter,
-> 1635             random_state=self.random_state))

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in evaluate_candidates(candidate_params, cv, more_results)
    807                                    (split_idx, (train, test)) in product(
    808                                    enumerate(candidate_params),
--> 809                                    enumerate(cv.split(X, y, groups))))
    810 
    811                 if len(out) < 1:

~\anaconda3\lib\site-packages\joblib\parallel.py in __call__(self, iterable)
   1015 
   1016             with self._backend.retrieval_context():
-> 1017                 self.retrieve()
   1018             # Make sure that we get a last message telling us we are done
   1019             elapsed_time = time.time() - self._start_time

~\anaconda3\lib\site-packages\joblib\parallel.py in retrieve(self)
    907             try:
    908                 if getattr(self._backend, 'supports_timeout', False):
--> 909                     self._output.extend(job.get(timeout=self.timeout))
    910                 else:
    911                     self._output.extend(job.get())

~\anaconda3\lib\site-packages\joblib\_parallel_backends.py in wrap_future_result(future, timeout)
    560         AsyncResults.get from multiprocessing."""
    561         try:
--> 562             return future.result(timeout=timeout)
    563         except LokyTimeoutError:
    564             raise TimeoutError()

~\anaconda3\lib\concurrent\futures\_base.py in result(self, timeout)
    433                 raise CancelledError()
    434             elif self._state == FINISHED:
--> 435                 return self.__get_result()
    436             else:
    437                 raise TimeoutError()

~\anaconda3\lib\concurrent\futures\_base.py in __get_result(self)
    382     def __get_result(self):
    383         if self._exception:
--> 384             raise self._exception
    385         else:
    386             return self._result

TypeError: Expected sequence or array-like, got <class 'sklearn.pipeline.Pipeline'>

当 verbose = 10 和 n_jobs = 1 时，会显示以下错误消息：

Fitting 5 folds for each of 2 candidates, totalling 10 fits
[CV 1/5; 1/2] START xgb_regressor__colsample_bytree=0.5, xgb_regressor__learning_rate=0.5, xgb_regressor__max_depth=5, xgb_regressor__n_estimators=5, xgb_regressor__subsample=0.5
C:\Users\Jan\anaconda3\lib\site-packages\sklearn\utils\validation.py:72: FutureWarning: Pass sample_weight=406477    1.0
393150    0.0
252885    0.0
260652    0.0
661256    0.0
         ... 
154663    0.0
398414    0.0
42890     0.0
640774    0.0
114446    0.0
Name: frequency, Length: 108482, dtype: float64 as keyword args. From version 1.0 (renaming of 0.25) passing these as positional arguments will result in an error
  "will result in an error", FutureWarning)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-84-74435f74c470> in <module>
----> 1 rscv.fit(X_train, y_train, xgb_regressor__sample_weight = X_train['Exposure'])
      2 #pipe_xgbr.fit(X_train, y_train)
      3 #X_train.describe(include = 'all')

~\anaconda3\lib\site-packages\sklearn\utils\validation.py in inner_f(*args, **kwargs)
     61             extra_args = len(args) - len(all_args)
     62             if extra_args <= 0:
---> 63                 return f(*args, **kwargs)
     64 
     65             # extra_args > 0

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in fit(self, X, y, groups, **fit_params)
    839                 return results
    840 
--> 841             self._run_search(evaluate_candidates)
    842 
    843             # multimetric is determined here because in the case of a callable

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in _run_search(self, evaluate_candidates)
   1633         evaluate_candidates(ParameterSampler(
   1634             self.param_distributions, self.n_iter,
-> 1635             random_state=self.random_state))

~\anaconda3\lib\site-packages\sklearn\model_selection\_search.py in evaluate_candidates(candidate_params, cv, more_results)
    807                                    (split_idx, (train, test)) in product(
    808                                    enumerate(candidate_params),
--> 809                                    enumerate(cv.split(X, y, groups))))
    810 
    811                 if len(out) < 1:

~\anaconda3\lib\site-packages\joblib\parallel.py in __call__(self, iterable)
   1002             # remaining jobs.
   1003             self._iterating = False
-> 1004             if self.dispatch_one_batch(iterator):
   1005                 self._iterating = self._original_iterator is not None
   1006 

~\anaconda3\lib\site-packages\joblib\parallel.py in dispatch_one_batch(self, iterator)
    833                 return False
    834             else:
--> 835                 self._dispatch(tasks)
    836                 return True
    837 

~\anaconda3\lib\site-packages\joblib\parallel.py in _dispatch(self, batch)
    752         with self._lock:
    753             job_idx = len(self._jobs)
--> 754             job = self._backend.apply_async(batch, callback=cb)
    755             # A job can complete so quickly than its callback is
    756             # called before we get here, causing self._jobs to

~\anaconda3\lib\site-packages\joblib\_parallel_backends.py in apply_async(self, func, callback)
    207     def apply_async(self, func, callback=None):
    208         """Schedule a func to be run"""
--> 209         result = ImmediateResult(func)
    210         if callback:
    211             callback(result)

~\anaconda3\lib\site-packages\joblib\_parallel_backends.py in __init__(self, batch)
    588         # Don't delay the application, to avoid keeping the input
    589         # arguments in memory
--> 590         self.results = batch()
    591 
    592     def get(self):

~\anaconda3\lib\site-packages\joblib\parallel.py in __call__(self)
    254         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    255             return [func(*args, **kwargs)
--> 256                     for func, args, kwargs in self.items]
    257 
    258     def __len__(self):

~\anaconda3\lib\site-packages\joblib\parallel.py in <listcomp>(.0)
    254         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    255             return [func(*args, **kwargs)
--> 256                     for func, args, kwargs in self.items]
    257 
    258     def __len__(self):

~\anaconda3\lib\site-packages\sklearn\utils\fixes.py in __call__(self, *args, **kwargs)
    220     def __call__(self, *args, **kwargs):
    221         with config_context(**self.config):
--> 222             return self.function(*args, **kwargs)

~\anaconda3\lib\site-packages\sklearn\model_selection\_validation.py in _fit_and_score(estimator, X, y, scorer, train, test, verbose, parameters, fit_params, return_train_score, return_parameters, return_n_test_samples, return_times, return_estimator, split_progress, candidate_progress, error_score)
    623 
    624         fit_time = time.time() - start_time
--> 625         test_scores = _score(estimator, X_test, y_test, scorer, error_score)
    626         score_time = time.time() - start_time - fit_time
    627         if return_train_score:

~\anaconda3\lib\site-packages\sklearn\model_selection\_validation.py in _score(estimator, X_test, y_test, scorer, error_score)
    685             scores = scorer(estimator, X_test)
    686         else:
--> 687             scores = scorer(estimator, X_test, y_test)
    688     except Exception:
    689         if error_score == 'raise':

~\anaconda3\lib\site-packages\sklearn\utils\validation.py in inner_f(*args, **kwargs)
     72                           "will result in an error", FutureWarning)
     73             kwargs.update(zip(sig.parameters, args))
---> 74             return f(**kwargs)
     75         return inner_f
     76 

~\anaconda3\lib\site-packages\sklearn\metrics\_regression.py in mean_squared_error(y_true, y_pred, sample_weight, multioutput, squared)
    334     """
    335     y_type, y_true, y_pred, multioutput = _check_reg_targets(
--> 336         y_true, y_pred, multioutput)
    337     check_consistent_length(y_true, y_pred, sample_weight)
    338     output_errors = np.average((y_true - y_pred) ** 2, axis=0,

~\anaconda3\lib\site-packages\sklearn\metrics\_regression.py in _check_reg_targets(y_true, y_pred, multioutput, dtype)
     86         the dtype argument passed to check_array.
     87     """
---> 88     check_consistent_length(y_true, y_pred)
     89     y_true = check_array(y_true, ensure_2d=False, dtype=dtype)
     90     y_pred = check_array(y_pred, ensure_2d=False, dtype=dtype)

~\anaconda3\lib\site-packages\sklearn\utils\validation.py in check_consistent_length(*arrays)
    314     """
    315 
--> 316     lengths = [_num_samples(X) for X in arrays if X is not None]
    317     uniques = np.unique(lengths)
    318     if len(uniques) > 1:

~\anaconda3\lib\site-packages\sklearn\utils\validation.py in <listcomp>(.0)
    314     """
    315 
--> 316     lengths = [_num_samples(X) for X in arrays if X is not None]
    317     uniques = np.unique(lengths)
    318     if len(uniques) > 1:

~\anaconda3\lib\site-packages\sklearn\utils\validation.py in _num_samples(x)
    247     if hasattr(x, 'fit') and callable(x.fit):
    248         # Don't get num_samples from an ensembles length!
--> 249         raise TypeError(message)
    250 
    251     if not hasattr(x, '__len__') and not hasattr(x, 'shape'):

TypeError: Expected sequence or array-like, got <class 'sklearn.pipeline.Pipeline'>

- Django

---

请问您能否提供代码中数据框 data_freq 的前几行，以便我们可以重现错误并更好地帮助您？ - TC Arlen

我添加了原数据框的前五行。希望这有所帮助。谢谢。 - Django

很高兴知道有一个参数错误分数。我已经将错误消息添加到原始帖子中。 - Django

哎呀，我还是看不出sklearn在_run_search部分中可能会看到期望序列/数组的管道。你能否尝试在没有搜索的情况下拟合管道？ - Ben Reiniger

pipe_xgbr.fit(X_train, y_train, xgb_regressor__sample_weight = X_train['Exposure']) 运行时没有抛出任何错误信息。 - Django

Answer 1

哇，那是一堆令人头痛的回溯信息，但我认为我终于找到问题所在了。你设置了scoring=mean_squared_error，但应该使用scoring="neg_mean_squared_error"。

mean_squared_error指标函数的签名为(y_true, y_pred, *, <kwargs>)，而通过使用字符串"neg_mean_squared_error"获取的得分器的签名为(estimator, X_test, y_test)。因此，在回溯中，你看到的是

--> 687             scores = scorer(estimator, X_test, y_test)

代码调用了mean_squared_error函数，并传入参数y_true=estimator、y_test=X_test和sample_weight=y_test（这是第一个关键字参数，因此会出现FutureWarning有关指定关键字参数的位置参数的警告）。深入跟踪发现，代码检查y_true和y_pred的形状是否兼容，但它把前者误认为是你的管道对象，最终导致错误信息的出现！

- Ben Reiniger

---

非常感谢@BenReiniger抽出时间来帮忙，将评分设置为'neg_mean_squared_error'就可以了。既然你提到了，我想我在某个地方读到过RandomizedSeachCV倾向于最大化得分函数。因此，您需要使用neg_mean_squared_error来最小化均方误差。 - Django

Answer 2

根据您的错误信息，KeyError: 'xgbr_regressor' 代码无法在管道中找到键为xgbr_regressor的对象。在您的管道中，已经定义了xgb_regressor:

pipe_xgbr = Pipeline(
  [('cf_trans', ct),
   ('ssc', StandardScaler(with_mean = False)),
   ('xgb_regressor', XGBRegressor())])

但是当你尝试去适配时，你使用了对 xgbr_regressor 的引用，这就是为什么会抛出 KeyError 错误：

rscv.fit(X_train, y_train, xgbr_regressor__sample_weight=X_train['Exposure'])

因此，您必须更改上面的行以交换 xgbr_regressor__sample_weight 为 xgb_regressor__sample_weight，这样应该就可以消除该错误。

- TC Arlen

---

好的，那是一个非常愚蠢的错误，对此我感到抱歉。不幸的是，这只解决了第一个错误。现在的错误消息与我发布的第二个错误消息相同（当省略sample_weight时）。 - Django