python SARIMX パラメータ最適化算出時のエラーについて

時系列分析でSARIMAXを使用しようとして
パラメータ最適算出しようとするとエラーがでて困っています。

このエラーの意味か　対処法について、どなたかアドバイスをお願いします。

データはアイスクリームの月次売上の様な、１２か月の周期性をもった、月間販売数です。

import optuna

s = 12

def objective(trial):
    # 各次数の探索範囲を設定します
    p = trial.suggest_int('p', 0, 3)
    d = trial.suggest_int('d', 0, 2)
    q = trial.suggest_int('q', 0, 3)
    P = trial.suggest_int('P', 0, 3)
    D = trial.suggest_int('D', 0, 2)
    Q = trial.suggest_int('Q', 0, 3)
    # ↑から選ばれた次数を用いてモデルを構築します
    model = sm.tsa.statespace.SARIMAX(raw_data_202003, order=(p, d, q), seasonal_order=(P, D, Q, s),
                                      enforce_stationarity=False, enforce_invertibility=False)
    fitting_result = model.fit()
    # BICの値を取得します
    bic = fitting_result.bic
    return bic

# 目的関数objectiveの値を最小化するようなパラメータを探索します
study = optuna.create_study(direction='minimize')
# 　試行回数を20回に指定します
study.optimize(objective,20)
trial_dataframe = study.trials_dataframe()
# BICが最小となるパラメータの組み合わせ
idx = trial_dataframe.value.idxmin()

でてくるエラーは、

TypeError                                 Traceback (most recent call last)
<ipython-input-65-a81496b1dd4a> in <module>()
     22 study = optuna.create_study(direction='minimize')
     23 # 　試行回数を20回に指定します
---> 24 study.optimize(objective,5)
     25 trial_dataframe = study.trials_dataframe()
     26 # BICが最小となるパラメータの組み合わせ

~\Anaconda3\lib\site-packages\optuna\study.py in optimize(self, func, n_trials, timeout, n_jobs, catch, callbacks, gc_after_trial)
    259             if n_jobs == 1:
    260                 self._optimize_sequential(func, n_trials, timeout, catch, callbacks,
--> 261                                           gc_after_trial)
    262             else:
    263                 self._optimize_parallel(func, n_trials, timeout, n_jobs, catch, callbacks,

~\Anaconda3\lib\site-packages\optuna\study.py in _optimize_sequential(self, func, n_trials, timeout, catch, callbacks, gc_after_trial)
    441                     break
    442 
--> 443             self._run_trial_and_callbacks(func, catch, callbacks, gc_after_trial)
    444 
    445     def _optimize_parallel(

~\Anaconda3\lib\site-packages\optuna\study.py in _run_trial_and_callbacks(self, func, catch, callbacks, gc_after_trial)
    518         # type: (...) -> None
    519 
--> 520         trial = self._run_trial(func, catch, gc_after_trial)
    521         if callbacks is not None:
    522             frozen_trial = self._storage.get_trial(trial._trial_id)

~\Anaconda3\lib\site-packages\optuna\study.py in _run_trial(self, func, catch, gc_after_trial)
    537 
    538         try:
--> 539             result = func(trial)
    540         except exceptions.TrialPruned as e:
    541             message = 'Setting status of trial#{} as {}. {}'.format(trial_number,

<ipython-input-65-a81496b1dd4a> in objective(trial)
     13     # ↑から選ばれた次数を用いてモデルを構築します
     14     model = sm.tsa.statespace.SARIMAX(raw_data_202003, order=(p, d, q), seasonal_order=(P, D, Q, s),
---> 15                                       enforce_stationarity=False, enforce_invertibility=False)
     16     fitting_result = model.fit()
     17     # BICの値を取得します

~\Anaconda3\lib\site-packages\statsmodels\tsa\statespace\sarimax.py in __init__(self, endog, exog, order, seasonal_order, trend, measurement_error, time_varying_regression, mle_regression, simple_differencing, enforce_stationarity, enforce_invertibility, hamilton_representation, **kwargs)
    497         # Initialize the statespace
    498         super(SARIMAX, self).__init__(
--> 499             endog, exog=exog, k_states=k_states, k_posdef=k_posdef, **kwargs
    500         )
    501 

~\Anaconda3\lib\site-packages\statsmodels\tsa\statespace\mlemodel.py in __init__(self, endog, k_states, exog, dates, freq, **kwargs)
    111         super(MLEModel, self).__init__(endog=endog, exog=exog,
    112                                        dates=dates, freq=freq,
--> 113                                        missing='none')
    114 
    115         # Store kwargs to recreate model

~\Anaconda3\lib\site-packages\statsmodels\tsa\base\tsa_model.py in __init__(self, endog, exog, dates, freq, missing, **kwargs)
     56 
     57         # Date handling in indexes
---> 58         self._init_dates(dates, freq)
     59 
     60     def _init_dates(self, dates=None, freq=None):

~\Anaconda3\lib\site-packages\statsmodels\tsa\base\tsa_model.py in _init_dates(self, dates, freq)
    189                 elif freq is not None and index.freq is None:
    190                     resampled_index = type(index)(
--> 191                         start=index[0], end=index[-1], freq=freq)
    192                     if not inferred_freq and not resampled_index.equals(index):
    193                         raise ValueError('The given frequency argument could'

TypeError: __new__() got an unexpected keyword argument 'start'