DOC: Improve regression doc strings

Improve regression doc strings
Fix small spelling errors

docs/source/spelling_wordlist.txt

@@ -2259,3 +2259,4 @@ statawriter
 Nonparameteric
 exceedance
 separatevar
+OaxacaResults

statsmodels/base/model.py

@@ -252,6 +252,15 @@ def initialize(self):
     def loglike(self, params):
         """
         Log-likelihood of model.
+
+        Parameters
+        ----------
+        params : ndarray
+            The model parameters used to compute the log-likelihood.
+
+        Notes
+        -----
+        Must be overridden by subclasses.
         """
         raise NotImplementedError
 
@@ -983,9 +992,18 @@ def __init__(self, model, params, **kwd):
         self.initialize(model, params, **kwd)
         self._data_attr = []
 
-    def initialize(self, model, params, **kwd):
+    def initialize(self, model, params, **kwargs):
         """
         Initialize (possibly re-initialize) a Results instance.
+
+        Parameters
+        ----------
+        model : Model
+            The model instance.
+        params : ndarray
+            The model parameters.
+        **kwargs
+            Any additional keyword arguments required to initialize the model.
         """
         self.params = params
         self.model = model
@@ -1016,7 +1034,7 @@ def predict(self, exog=None, transform=True, *args, **kwargs):
 
         Returns
         -------
-        prediction : {ndarray, Series, DataFrame}
+        array_like
             See self.model.predict.
 
         Notes

statsmodels/base/wrapper.py

@@ -1,8 +1,9 @@
-import inspect
-import functools
-
 from statsmodels.compat.python import iteritems
 
+import functools
+import inspect
+from textwrap import dedent
+
 
 class ResultsWrapper(object):
     """
@@ -43,11 +44,11 @@ def __getattribute__(self, attr):
         return obj
 
     def __getstate__(self):
-        #print 'pickling wrapper', self.__dict__
+        # print 'pickling wrapper', self.__dict__
         return self.__dict__
 
     def __setstate__(self, dict_):
-        #print 'unpickling wrapper', dict_
+        # print 'unpickling wrapper', dict_
         self.__dict__.update(dict_)
 
     def save(self, fname, remove_data=False):
@@ -111,7 +112,8 @@ def wrapper(self, *args, **kwargs):
     sig = inspect.signature(func)
     formatted = str(sig)
 
-    wrapper.__doc__ = "%s%s\n%s" % (func.__name__, formatted, wrapper.__doc__)
+    doc = dedent(wrapper.__doc__) if wrapper.__doc__ else ''
+    wrapper.__doc__ = "\n%s%s\n%s" % (func.__name__, formatted, doc)
 
     return wrapper
 

statsmodels/graphics/regressionplots.py

@@ -759,7 +759,7 @@ def abline_plot(intercept=None, slope=None, horiz=None, vert=None,
 
     Returns
     -------
-    fig : Figure
+    Figure
         The figure given by `ax.figure` or a new instance.
 
     Examples
@@ -779,7 +779,6 @@ def abline_plot(intercept=None, slope=None, horiz=None, vert=None,
     >>> plt.show()
 
     .. plot:: plots/graphics_regression_abline.py
-
     """
     if ax is not None:  # get axis limits first thing, do not change these
         x = ax.get_xlim()

statsmodels/regression/_prediction.py

@@ -14,7 +14,23 @@
 # this is similar to ContrastResults after t_test, copied and adjusted
 class PredictionResults(object):
     """
-    Results class for predictions
+    Results class for predictions.
+
+    Parameters
+    ----------
+    predicted_mean : ndarray
+        The array containing the prediction means.
+    var_pred_mean : ndarray
+        The array of the variance of the prediction means.
+    var_resid : ndarray
+        The array of residual variances.
+    df : int
+        The degree of freedom used if dist is 't'.
+    dist : {'norm', 't', object}
+        Either a string for the normal or t distribution or another object
+        that exposes a `ppf` method.
+    row_labels : list[str]
+        Row labels used in summary frame.
     """
 
     def __init__(self, predicted_mean, var_pred_mean, var_resid,
@@ -61,7 +77,6 @@ def conf_int(self, obs=False, alpha=0.05):
         ci : ndarray, (k_constraints, 2)
             The array has the lower and the upper limit of the confidence
             interval in the columns.
-
         """
 
         se = self.se_obs if obs else self.se_mean
@@ -75,7 +90,7 @@ def summary_frame(self, what='all', alpha=0.05):
         # TODO: finish and cleanup
         import pandas as pd
         from collections import OrderedDict
-        ci_obs = self.conf_int(alpha=alpha, obs=True) # need to split
+        ci_obs = self.conf_int(alpha=alpha, obs=True)  # need to split
         ci_mean = self.conf_int(alpha=alpha, obs=False)
         to_include = OrderedDict()
         to_include['mean'] = self.predicted_mean
@@ -86,10 +101,10 @@ def summary_frame(self, what='all', alpha=0.05):
         to_include['obs_ci_upper'] = ci_obs[:, 1]
 
         self.table = to_include
-        #OrderedDict does not work to preserve sequence
+        # OrderedDict does not work to preserve sequence
         # pandas dict does not handle 2d_array
-        #data = np.column_stack(list(to_include.values()))
-        #names = ....
+        # data = np.column_stack(list(to_include.values()))
+        # names = ....
         res = pd.DataFrame(to_include, index=self.row_labels,
                            columns=to_include.keys())
         return res
@@ -157,11 +172,9 @@ def get_prediction(self, exog=None, transform=True, weights=None,
     if weights is not None:
         weights = np.asarray(weights)
         if (weights.size > 1 and
-           (weights.ndim != 1 or weights.shape[0] == exog.shape[1])):
+                (weights.ndim != 1 or weights.shape[0] == exog.shape[1])):
             raise ValueError('weights has wrong shape')
 
-    ### end
-
     if pred_kwds is None:
         pred_kwds = {}
     predicted_mean = self.model.predict(self.params, exog, **pred_kwds)