Fixed issue with a_vector.py that was affecting class method pm.summa…

…ry mentioned in #11. Other minor cleanup removing Bayesian artifacts from master branch metalog.py file. Going to rebuild and publish to pypi

pymetalog/a_vector.py

@@ -153,8 +153,9 @@ def a_vector_OLS_and_LP(m_dict,
             try:
                 temp = np.dot(np.dot(np.linalg.inv(np.dot(Y.T, Y) + alpha*eye), Y.T), z)
             except:
-                temp = a_vector_LP(m_dict, term_limit=i, term_lower_bound=i, diff_error=diff_error, diff_step=diff_step)
                 # use LP solver if OLS breaks
+                temp = a_vector_LP(m_dict, term_limit=i, term_lower_bound=i, diff_error=diff_error, diff_step=diff_step)
+                methodFit = 'Linear Program'
         if fit_method == 'OLS':
             try:
                 temp = np.dot(np.dot(np.linalg.inv(np.dot(Y.T, Y) + alpha*eye), Y.T), z)
@@ -164,10 +165,11 @@ def a_vector_OLS_and_LP(m_dict,
             temp = a_vector_LP(m_dict, term_limit=i, term_lower_bound=i, diff_error=diff_error, diff_step=diff_step)
             methodFit = 'Linear Program'
 
-        temp = np.append(temp, np.zeros(term_limit-i))
         if fit_method == 'MLE':
             temp = a_vector_MLE(temp, y, i, m_dict, bounds, boundedness)
 
+        temp = np.append(temp, np.zeros(term_limit-i))
+
         # build a y vector for smaller data sets
         if len(z) < 100:
             y2 = np.linspace(step_len, 1 - step_len, int((1 - step_len) / step_len))

pymetalog/metalog.py

@@ -36,7 +36,6 @@ class metalog():
         fit_method (:obj: `str`): String type of metalog fit method ('any' | 'OLS' | 'LP' | 'MLE').
         penalty (:obj:`str`): Used to specify the norm used in the regularization.
         alpha (:obj:`float`): Regularization term to add to OLS fit.
-        save_data (:obj:`bool`, optional): T/F to save copy of input data `x` with metalog object.
 
         output_dict (:obj:`dict` with keys ['params', 'dataValues', 'Y', 'A', 'M', 'Validation']).
             - output_dict['params'] (:obj:`dict`):
@@ -92,7 +91,7 @@ class metalog():
         append_zvector(`bounds`, `boundedness`) -> df_x: (:obj:`pandas.DataFrame` with columns ['x','probs','z'] of type numeric)
         
     """
-    def __init__(self, x, bounds=[0,1], boundedness='u', term_limit=13, term_lower_bound=2, step_len=.01, probs=None, fit_method='any', penalty=None, alpha=0., save_data = False):
+    def __init__(self, x, bounds=[0,1], boundedness='u', term_limit=13, term_lower_bound=2, step_len=.01, probs=None, fit_method='any', penalty=None, alpha=0.):
         """Fits a metalog distribution using the input array `x`.
 
         Args:
@@ -150,12 +149,6 @@ def __init__(self, x, bounds=[0,1], boundedness='u', term_limit=13, term_lower_b
                 - should be set in conjunction with `penalty` parameter
                 - Default: 0. (no regularization, OLS)
 
-            save_data (:obj:`bool`, optional): Saves copy of input data `x` with metalog object.
-                - strictly boolean (True, False)
-                - saves a copy of data used to fit metalog object
-                    * Used for Bayesian updating
-                - Default: False
-
         Raises:
             TypeError: 'Input x must be an array or pandas Series'
             TypeError: 'Input x must be an array of allowable types: int, float, numpy.int64, or numpy.float64'
@@ -185,7 +178,6 @@ def __init__(self, x, bounds=[0,1], boundedness='u', term_limit=13, term_lower_b
             ValueError: 'fit_method can only be values OLS, LP, any, or MLE'
             ValueError: 'penalty can only be values l2 or None'
             ValueError: 'alpha must only be a float >= 0.'
-            ValueError: 'save_data must only be boolean.'
 
         Example:
 
@@ -215,7 +207,6 @@ def __init__(self, x, bounds=[0,1], boundedness='u', term_limit=13, term_lower_b
         self.fit_method = fit_method
         self.penalty = penalty
         self.nobs = len(x)
-        self.save_data = save_data
 
         if penalty == None:
             alpha = 0.
@@ -436,18 +427,6 @@ def alpha(self, a):
             raise ValueError('alpha must only be a float >= 0.')
         self._alpha = a
 
-    @property
-    def save_data(self):
-        """save_data (:obj:`bool`, optional): Saves copy of input data `x` with metalog object."""
-
-        return self._save_data
-
-    @save_data.setter
-    def save_data(self, sd):
-        if not type(sd) is bool:
-            raise ValueError('save_data must only be boolean.')
-        self._save_data = sd
-
     def get_params(self):
         """Sets the `params` key (dict) of `output_dict` object prior to input to `a_vector_OLS_and_LP` method.
             - Uses metalog attributes to set keys
@@ -465,7 +444,6 @@ def get_params(self):
         params['step_len'] = self.step_len
         params['fit_method'] = self.fit_method
         params['nobs'] = self.nobs
-        params['save_data'] = self.save_data
 
         return params