Missing DataΒΆ

All of the models can handle missing data. For performance reasons, the default is not to do any checking for missing data. If, however, you would like for missing data to be handled internally, you can do so by using the missing keyword argument. The default is to do nothing

In [1]: import statsmodels.api as sm

In [2]: data = sm.datasets.longley.load()

In [3]: data.exog = sm.add_constant(data.exog)

# add in some missing data
In [4]: missing_idx = np.array([False] * len(data.endog))

In [5]: missing_idx[[4, 10, 15]] = True

In [6]: data.endog[missing_idx] = np.nan

In [7]: ols_model = sm.OLS(data.endog, data.exog)

In [8]: ols_fit = ols_model.fit()

In [9]: print(ols_fit.params)
[ nan  nan  nan  nan  nan  nan  nan]

This silently fails and all of the model parameters are NaN, which is probably not what you expected. If you are not sure whether or not you have missing data you can use missing = ‘raise’. This will raise a MissingDataError during model instantiation if missing data is present so that you know something was wrong in your input data.

In [10]: ols_model = sm.OLS(data.endog, data.exog, missing='raise')
---------------------------------------------------------------------------
MissingDataError                          Traceback (most recent call last)
<ipython-input-10-6b74d5399bc3> in <module>()
----> 1 ols_model = sm.OLS(data.endog, data.exog, missing='raise')

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/regression/linear_model.py in __init__(self, endog, exog, missing, hasconst, **kwargs)
    815                  **kwargs):
    816         super(OLS, self).__init__(endog, exog, missing=missing,
--> 817                                   hasconst=hasconst, **kwargs)
    818         if "weights" in self._init_keys:
    819             self._init_keys.remove("weights")

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/regression/linear_model.py in __init__(self, endog, exog, weights, missing, hasconst, **kwargs)
    661             weights = weights.squeeze()
    662         super(WLS, self).__init__(endog, exog, missing=missing,
--> 663                                   weights=weights, hasconst=hasconst, **kwargs)
    664         nobs = self.exog.shape[0]
    665         weights = self.weights

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/regression/linear_model.py in __init__(self, endog, exog, **kwargs)
    177     """
    178     def __init__(self, endog, exog, **kwargs):
--> 179         super(RegressionModel, self).__init__(endog, exog, **kwargs)
    180         self._data_attr.extend(['pinv_wexog', 'wendog', 'wexog', 'weights'])
    181 

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/base/model.py in __init__(self, endog, exog, **kwargs)
    210 
    211     def __init__(self, endog, exog=None, **kwargs):
--> 212         super(LikelihoodModel, self).__init__(endog, exog, **kwargs)
    213         self.initialize()
    214 

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/base/model.py in __init__(self, endog, exog, **kwargs)
     62         hasconst = kwargs.pop('hasconst', None)
     63         self.data = self._handle_data(endog, exog, missing, hasconst,
---> 64                                       **kwargs)
     65         self.k_constant = self.data.k_constant
     66         self.exog = self.data.exog

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/base/model.py in _handle_data(self, endog, exog, missing, hasconst, **kwargs)
     85 
     86     def _handle_data(self, endog, exog, missing, hasconst, **kwargs):
---> 87         data = handle_data(endog, exog, missing, hasconst, **kwargs)
     88         # kwargs arrays could have changed, easier to just attach here
     89         for key in kwargs:

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/base/data.py in handle_data(endog, exog, missing, hasconst, **kwargs)
    631     klass = handle_data_class_factory(endog, exog)
    632     return klass(endog, exog=exog, missing=missing, hasconst=hasconst,
--> 633                  **kwargs)

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/base/data.py in __init__(self, endog, exog, missing, hasconst, **kwargs)
     63         if missing != 'none':
     64             arrays, nan_idx = self.handle_missing(endog, exog, missing,
---> 65                                                   **kwargs)
     66             self.missing_row_idx = nan_idx
     67             self.__dict__.update(arrays)  # attach all the data arrays

/build/statsmodels-0.9.0/.pybuild/pythonX.Y_3.5/build/statsmodels/base/data.py in handle_missing(cls, endog, exog, missing, **kwargs)
    279 
    280         elif missing == 'raise':
--> 281             raise MissingDataError("NaNs were encountered in the data")
    282 
    283         elif missing == 'drop':

MissingDataError: NaNs were encountered in the data

If you want statsmodels to handle the missing data by dropping the observations, use missing = ‘drop’.

In [11]: ols_model = sm.OLS(data.endog, data.exog, missing='drop')

We are considering adding a configuration framework so that you can set the option with a global setting.