Merge 0f24c10 into 0529da7

statsmodels/base/model.py

@@ -596,7 +596,7 @@ def score(self, params):
         kwds.setdefault('centered', True)
         return approx_fprime(params, self.loglike, **kwds).ravel()
 
-    def jac(self, params, **kwds):
+    def score_obs(self, params, **kwds):
         '''
         Jacobian/Gradient of log-likelihood evaluated at params for each
         observation.
@@ -605,6 +605,9 @@ def jac(self, params, **kwds):
         kwds.setdefault('centered', True)
         return approx_fprime(params, self.loglikeobs, **kwds)
 
+    jac = np.deprecate(score_obs, 'jac', 'score_obs', "Use score_obs method."
+                       " jac will be removed in 0.7.")
+
     def hessian(self, params):
         '''
         Hessian of log-likelihood evaluated at params
@@ -1609,10 +1612,14 @@ def bic(self):
         return -2 * self.llf + np.log(self.nobs) * (self.df_modelwc)
 
     @cache_readonly
-    def jacv(self):
+    def score_obsv(self):
         '''cached Jacobian of log-likelihood
         '''
-        return self.model.jac(self.params)
+        return self.model.score_obs(self.params)
+
+    jacv = np.deprecate(score_obsv, 'jacv', 'score_obsv',
+                        "Use score_obsv attribute."
+                       " jacv will be removed in 0.7.")
 
     @cache_readonly
     def hessv(self):
@@ -1631,7 +1638,7 @@ def covjac(self):
         ##      raise ValueError('need to call fit first')
         ##      #self.fit()
         ##  self.jacv = jacv = self.jac(self._results.params)
-        jacv = self.jacv
+        jacv = self.score_obsv
         return np.linalg.inv(np.dot(jacv.T, jacv))
 
     @cache_readonly
@@ -1642,11 +1649,10 @@ def covjhj(self):
 
         name should be covhjh
         '''
-        jacv = self.jacv
-        ##  hessv = self.hessv
-        ##  hessinv = np.linalg.inv(hessv)
-        ##  self.hessinv = hessinv
-        hessinv = self.cov_params()
+        jacv = self.score_obsv
+        hessv = self.hessv
+        hessinv = np.linalg.inv(hessv)
+        ##  self.hessinv = hessin = self.cov_params()
         return np.dot(hessinv, np.dot(np.dot(jacv.T, jacv), hessinv))
 
     @cache_readonly

statsmodels/discrete/discrete_model.py

@@ -1056,7 +1056,7 @@ def score(self, params):
         L = np.exp(np.dot(X,params) + offset + exposure)
         return np.dot(self.endog - L, X)
 
-    def jac(self, params):
+    def score_obs(self, params):
         """
         Poisson model Jacobian of the log-likelihood for each observation
 
@@ -1086,6 +1086,9 @@ def jac(self, params):
         L = np.exp(np.dot(X,params) + offset + exposure)
         return (self.endog - L)[:,None] * X
 
+    jac = np.deprecate(score_obs, 'jac', 'score_obs', "Use score_obs method."
+                       " jac will be removed in 0.7")
+
     def hessian(self, params):
         """
         Poisson model Hessian matrix of the loglikelihood
@@ -1257,7 +1260,7 @@ def score(self, params):
         L = self.cdf(np.dot(X,params))
         return np.dot(y - L,X)
 
-    def jac(self, params):
+    def score_obs(self, params):
         """
         Logit model Jacobian of the log-likelihood for each observation
 
@@ -1285,6 +1288,9 @@ def jac(self, params):
         L = self.cdf(np.dot(X, params))
         return (y - L)[:,None] * X
 
+    jac = np.deprecate(score_obs, 'jac', 'score_obs', "Use score_obs method."
+                       " jac will be removed in 0.7")
+
     def hessian(self, params):
         """
         Logit model Hessian matrix of the log-likelihood
@@ -1465,7 +1471,7 @@ def score(self, params):
         L = q*self.pdf(q*XB)/np.clip(self.cdf(q*XB), FLOAT_EPS, 1 - FLOAT_EPS)
         return np.dot(L,X)
 
-    def jac(self, params):
+    def score_obs(self, params):
         """
         Probit model Jacobian for each observation
 
@@ -1497,6 +1503,9 @@ def jac(self, params):
         L = q*self.pdf(q*XB)/np.clip(self.cdf(q*XB), FLOAT_EPS, 1 - FLOAT_EPS)
         return L[:,None] * X
 
+    jac = np.deprecate(score_obs, 'jac', 'score_obs', "Use score_obs method."
+                       " jac will be removed in 0.7")
+
     def hessian(self, params):
         """
         Probit model Hessian matrix of the log-likelihood
@@ -1710,7 +1719,7 @@ def loglike_and_score(self, params):
         score_array = np.dot(firstterm.T, self.exog).flatten()
         return loglike_value, score_array
 
-    def jac(self, params):
+    def score_obs(self, params):
         """
         Jacobian matrix for multinomial logit model log-likelihood
 
@@ -1741,6 +1750,9 @@ def jac(self, params):
         #NOTE: might need to switch terms if params is reshaped
         return (firstterm[:,:,None] * self.exog[:,None,:]).reshape(self.exog.shape[0], -1)
 
+    jac = np.deprecate(score_obs, 'jac', 'score_obs', "Use score_obs method."
+                       " jac will be removed in 0.7")
+
     def hessian(self, params):
         """
         Multinomial logit Hessian matrix of the log-likelihood
@@ -2180,10 +2192,13 @@ def _hessian_nb2(self, params):
         return hess_arr
 
     #TODO: replace this with analytic where is it used?
-    def jac(self, params):
+    def score_obs(self, params):
         sc = approx_fprime_cs(params, self.loglikeobs)
         return sc
 
+    jac = np.deprecate(score_obs, 'jac', 'score_obs', "Use score_obs method."
+                       " jac will be removed in 0.7")
+
     def fit(self, start_params=None, method='bfgs', maxiter=35,
             full_output=1, disp=1, callback=None,
             cov_type='nonrobust', cov_kwds=None, use_t=None, **kwargs):

statsmodels/discrete/tests/test_discrete.py

@@ -111,7 +111,7 @@ def test_loglikeobs(self):
 
     def test_jac(self):
         #basic cross check
-        jacsum = self.res1.model.jac(self.res1.params).sum(0)
+        jacsum = self.res1.model.score_obs(self.res1.params).sum(0)
         score = self.res1.model.score(self.res1.params)
         assert_almost_equal(jacsum, score, DECIMAL_9) #Poisson has low precision ?
 

statsmodels/examples/ex_generic_mle.py

@@ -373,11 +373,11 @@ def loglikeobs(self, params):
 Is scale a misnomer, actually scale squared, i.e. variance of error term ?
 '''
 
-print(res_norm3.model.jac(res_norm3.params).shape)
+print(res_norm3.model.score_obs(res_norm3.params).shape)
 
-jac = res_norm3.model.jac(res_norm3.params)
+jac = res_norm3.model.score_obs(res_norm3.params)
 print(np.sqrt(np.diag(np.dot(jac.T, jac)))/start_params)
-jac2 = res_norm3.model.jac(res_norm3.params, centered=True)
+jac2 = res_norm3.model.score_obs(res_norm3.params, centered=True)
 
 print(np.sqrt(np.diag(np.linalg.inv(np.dot(jac.T, jac)))))
 print(res_norm3.bse)

statsmodels/genmod/tests/test_glm.py

@@ -130,7 +130,7 @@ def test_compare_discrete(self):
 
         assert_allclose(res1.llf, resd.llf, rtol=1e-10)
         score_obs1 = res1.model.score_obs(res1.params)
-        score_obsd = resd.model.jac(resd.params)
+        score_obsd = resd.model.score_obs(resd.params)
         assert_allclose(score_obs1, score_obsd, rtol=1e-10)
 
         # score