Add return self

Author: yuta-nakahara

bayesml/autoregressive/_autoregressive.py

@@ -142,6 +142,7 @@ def set_h_params(self,h_mu_vec,h_lambda_mat,h_alpha,h_beta):
         if self.degree+1 != self.theta_vec.shape[0]:
             self.theta_vec = np.zeros(self.degree+1)
             warnings.warn("theta_vec is reinitialized to [0.0, 0.0, ..., 0.0] because dimension of theta_vec and h_parms are mismatched.", ParameterFormatWarning)
+        return self
 
     def get_h_params(self):
         """Get the hyperparameters of the prior distribution.
@@ -163,6 +164,7 @@ def gen_params(self):
         """
         self.tau = self.rng.gamma(shape=self.h_alpha,scale=1.0/self.h_beta)
         self.theta_vec = self.rng.multivariate_normal(mean=self.h_mu_vec,cov=np.linalg.inv(self.tau*self.h_lambda_mat))
+        return self
 
     def set_params(self,theta_vec, tau):
         """Set the parameter of the sthocastic data generative model.
@@ -184,6 +186,7 @@ def set_params(self,theta_vec, tau):
         if self.degree+1 != self.h_lambda_mat.shape[0]:
             self.h_lambda_mat = np.identity(self.degree+1)
             warnings.warn("h_lambda_mat is reinitialized to the identity matrix because dimension of h_lambda_mat and theta_vec are mismatched.", ParameterFormatWarning)
+        return self
 
     def get_params(self):
         """Get the parameter of the sthocastic data generative model.
@@ -406,6 +409,7 @@ def set_h0_params(self,h0_mu_vec,h0_lambda_mat,h0_alpha,h0_beta):
                     "dimensions of h0_mu_vec and h0_lambda_mat must be the same"))
 
         self.reset_hn_params()
+        return self
 
     def get_h0_params(self):
         """Get the initial values of the hyperparameters of the posterior distribution.
@@ -460,6 +464,7 @@ def set_hn_params(self,hn_mu_vec,hn_lambda_mat,hn_alpha,hn_beta):
                 raise(ParameterFormatError(
                     "dimensions of hn_mu_vec and hn_lambda_mat must be the same"))
         self.calc_pred_dist(np.zeros(self.degree))
+        return self
 
     def overwrite_h0_params(self):
         """Overwrite the initial values of the hyperparameters of the posterior distribution by the learned values.
@@ -474,6 +479,7 @@ def overwrite_h0_params(self):
         self.h0_beta = self.hn_beta
 
         self.calc_pred_dist(np.zeros(self.degree))
+        return self
 
     def reset_hn_params(self):
         """Reset the hyperparameters of the posterior distribution to their initial values.
@@ -488,6 +494,7 @@ def reset_hn_params(self):
         self.hn_beta = self.h0_beta
 
         self.calc_pred_dist(np.zeros(self.degree))
+        return self
 
     def update_posterior(self,x,padding=None):
         """Update the hyperparameters of the posterior distribution using traning data.
@@ -527,6 +534,7 @@ def update_posterior(self,x,padding=None):
             self.hn_alpha += (x.shape[0]-self.degree) / 2.0
             self.hn_beta += (-self.hn_mu_vec[np.newaxis,:] @ self.hn_lambda_mat @ self.hn_mu_vec[:,np.newaxis] 
                             + x[self.degree:] @ x[self.degree:] + mu_tmp[np.newaxis,:] @ lambda_tmp @ mu_tmp[:,np.newaxis])[0,0] / 2.0
+        return self
 
     def estimate_params(self,loss="squared"):
         """Estimate the parameter of the stochastic data generative model under the given criterion.
@@ -647,6 +655,7 @@ def calc_pred_dist(self,x):
         self.p_m = self.hn_mu_vec @ _explanatory_vec
         self.p_lambda = self.hn_alpha / self.hn_beta / (1.0 + _explanatory_vec @ np.linalg.solve(self.hn_lambda_mat,_explanatory_vec))
         self.p_nu = 2.0 * self.hn_alpha
+        return self
 
     def make_prediction(self,loss="squared"):
         """Predict a new data point under the given criterion.

bayesml/base.py

@@ -58,7 +58,7 @@ def load_h_params(self,filename):
             tmp_h_params = pickle.load(f)
         if type(tmp_h_params) is dict:
             self.set_h_params(*tmp_h_params.values())
-            return
+            return self
 
         raise(ParameterFormatError(
             filename+" must be a pickled python dictionary obtained by "
@@ -120,7 +120,7 @@ def load_params(self,filename):
             params = pickle.load(f)
         if type(params) is dict:
             self.set_params(*params.values())
-            return
+            return self
 
         raise(ParameterFormatError(filename+" must be a pickled python dictionary obtained by ``GenModel.save_params()``"))
 
@@ -189,7 +189,7 @@ def load_h0_params(self,filename):
             tmp_h_params = pickle.load(f)
         if type(tmp_h_params) is dict:
             self.set_h0_params(*tmp_h_params.values())
-            return
+            return self
 
         raise(ParameterFormatError(
             filename+" must be a pickled python dictionary obtained by "
@@ -249,7 +249,7 @@ def load_hn_params(self,filename):
             tmp_h_params = pickle.load(f)
         if type(tmp_h_params) is dict:
             self.set_hn_params(*tmp_h_params.values())
-            return
+            return self
 
         raise(ParameterFormatError(
             filename+" must be a pickled python dictionary obtained by "
@@ -264,6 +264,7 @@ def reset_hn_params(self):
         Note that the parameters of the predictive distribution are also calculated from them.
         """
         self.set_hn_params(*self.get_h0_params().values())
+        return self
 
     def overwrite_h0_params(self):
         """Overwrite the initial values of the hyperparameters of the posterior distribution by the learned values.
@@ -272,6 +273,7 @@ def overwrite_h0_params(self):
         Note that the parameters of the predictive distribution are also calculated from them.
         """
         self.set_h0_params(*self.get_hn_params().values())
+        return self
 
     @abstractmethod
     def update_posterior(self):

bayesml/bernoulli/_bernoulli.py

@@ -45,6 +45,7 @@ def set_h_params(self,h_alpha,h_beta):
         """
         self.h_alpha = _check.pos_float(h_alpha,'h_alpha',ParameterFormatError)
         self.h_beta = _check.pos_float(h_beta,'h_beta',ParameterFormatError)
+        return self
 
     def get_h_params(self):
         """Get the hyperparameters of the prior distribution.
@@ -63,7 +64,8 @@ def gen_params(self):
         The generated vaule is set at ``self.theta``.
         """
         self.theta = self.rng.beta(self.h_alpha,self.h_beta)
-        
+        return self
+
     def set_params(self,theta):
         """Set the parameter of the sthocastic data generative model.
 
@@ -73,6 +75,7 @@ def set_params(self,theta):
             a real number :math:`\theta \in [0, 1]`
         """
         self.theta = _check.float_in_closed01(theta,'theta',ParameterFormatError)
+        return self
 
     def get_params(self):
         """Get the parameter of the sthocastic data generative model.
@@ -203,6 +206,7 @@ def set_h0_params(self,h0_alpha,h0_beta):
         self.h0_alpha = _check.pos_float(h0_alpha,'h0_alpha',ParameterFormatError)
         self.h0_beta = _check.pos_float(h0_beta,'h0_beta',ParameterFormatError)
         self.reset_hn_params()
+        return self
 
     def get_h0_params(self):
         """Get the initial values of the hyperparameters of the posterior distribution.
@@ -228,6 +232,7 @@ def set_hn_params(self,hn_alpha,hn_beta):
         self.hn_alpha = _check.pos_float(hn_alpha,'hn_alpha',ParameterFormatError)
         self.hn_beta = _check.pos_float(hn_beta,'hn_beta',ParameterFormatError)
         self.calc_pred_dist()
+        return self
 
     def get_hn_params(self):
         """Get the hyperparameters of the posterior distribution.
@@ -249,6 +254,7 @@ def reset_hn_params(self):
         self.hn_alpha = self.h0_alpha
         self.hn_beta = self.h0_beta
         self.calc_pred_dist()
+        return self
 
     def overwrite_h0_params(self):
         """Overwrite the initial values of the hyperparameters of the posterior distribution by the learned values.
@@ -259,6 +265,7 @@ def overwrite_h0_params(self):
         self.h0_alpha = self.hn_alpha
         self.h0_beta = self.hn_beta
         self.calc_pred_dist()
+        return self
 
     def update_posterior(self,x):
         """Update the hyperparameters of the posterior distribution using traning data.
@@ -271,6 +278,7 @@ def update_posterior(self,x):
         _check.ints_of_01(x,'x',DataFormatError)
         self.hn_alpha += np.sum(x==1)
         self.hn_beta += np.sum(x==0)
+        return self
 
     def estimate_params(self,loss="squared",dict_out=False):
         """Estimate the parameter of the stochastic data generative model under the given criterion.
@@ -385,6 +393,7 @@ def get_p_params(self):
     def calc_pred_dist(self):
         """Calculate the parameters of the predictive distribution."""
         self.p_theta = self.hn_alpha / (self.hn_alpha + self.hn_beta)
+        return self
 
     def make_prediction(self,loss="squared"):
         """Predict a new data point under the given criterion.

bayesml/categorical/_categorical.py

@@ -96,6 +96,7 @@ def set_h_params(self,h_alpha_vec):
         if self.degree != self.theta_vec.shape[0]:
             self.theta_vec = np.ones(self.degree) / self.degree
             warnings.warn("theta_vec is reinitialized to [1.0/self.degree, 1.0/self.degree, ..., 1.0/self.degree] because dimension of theta_vec and h_alpha_vec are mismatched.", ParameterFormatWarning)
+        return self
 
     def get_h_params(self):
         """Get the hyperparameters of the prior distribution.
@@ -113,6 +114,7 @@ def gen_params(self):
         The generated vaule is set at ``self.theta_vec``.
         """
         self.theta_vec[:] = self.rng.dirichlet(self.h_alpha_vec)
+        return self
 
     def set_params(self, theta_vec):
         """Set the parameter of the sthocastic data generative model.
@@ -128,6 +130,7 @@ def set_params(self, theta_vec):
         if self.degree != self.h_alpha_vec.shape[0]:
             self.h_alpha_vec = np.ones(self.degree) / 2.0
             warnings.warn("h_alpha_vec is reinitialized to [1/2, 1/2, ..., 1/2] because dimension of h_m_vec and mu_vec are mismatched.", ParameterFormatWarning)
+        return self
 
     def get_params(self):
         """Get the parameter of the sthocastic data generative model.
@@ -289,6 +292,7 @@ def set_h0_params(self,h0_alpha_vec):
         self.h0_alpha_vec = _check.pos_float_vec(h0_alpha_vec,'h0_alpha_vec',ParameterFormatError)
         self.degree = self.h0_alpha_vec.shape[0]
         self.reset_hn_params()
+        return self
 
     def get_h0_params(self):
         """Get the initial values of the hyperparameters of the posterior distribution.
@@ -311,6 +315,7 @@ def set_hn_params(self,hn_alpha_vec):
         self.hn_alpha_vec = _check.pos_float_vec(hn_alpha_vec,'hn_alpha_vec',ParameterFormatError)
         self.degree = self.hn_alpha_vec.shape[0]
         self.calc_pred_dist()
+        return self
 
     def get_hn_params(self):
         """Get the hyperparameters of the posterior distribution.
@@ -330,6 +335,7 @@ def reset_hn_params(self):
         """
         self.hn_alpha_vec = np.copy(self.h0_alpha_vec)
         self.calc_pred_dist()
+        return self
 
     def overwrite_h0_params(self):
         """Overwrite the initial value of the hyperparameter of the posterior distribution by the learned values.
@@ -339,6 +345,7 @@ def overwrite_h0_params(self):
         """
         self.h0_alpha_vec = np.copy(self.hn_alpha_vec)
         self.calc_pred_dist()
+        return self
 
     def update_posterior(self, x):
         """Update the hyperparameters of the posterior distribution using traning data.
@@ -355,6 +362,7 @@ def update_posterior(self, x):
 
         for k in range(self.degree):
             self.hn_alpha_vec[k] += x[:,k].sum()
+        return self
 
     def estimate_params(self, loss="squared",dict_out=False):
         """Estimate the parameter of the stochastic data generative model under the given criterion.
@@ -471,6 +479,7 @@ def get_p_params(self):
     def calc_pred_dist(self):
         """Calculate the parameters of the predictive distribution."""
         self.p_theta_vec = self.hn_alpha_vec / self.hn_alpha_vec.sum()
+        return self
 
     def make_prediction(self, loss="squared"):
         """Predict a new data point under the given criterion.

bayesml/contexttree/_contexttree.py

@@ -295,6 +295,7 @@ def set_h_params(self,
                 ))
             self.h_root = _Node(0,self.c_k)
             self._set_h_params_recursion(self.h_root,h_root)
+        return self
 
     def get_h_params(self):
         """Get the hyperparameters of the prior distribution.
@@ -324,7 +325,8 @@ def gen_params(self,tree_fix=False):
             self._gen_params_recursion_tree_fix(self.root,self.h_root)
         else:
             self._gen_params_recursion(self.root,self.h_root)
-    
+        return self
+
     def set_params(self,root=None):
         """Set the parameter of the sthocastic data generative model.
 
@@ -339,6 +341,7 @@ def set_params(self,root=None):
                     "root must be an instance of metatree._Node"
                 ))
             self._set_params_recursion(self.root,root)
+        return self
 
     def get_params(self):
         """Get the parameter of the sthocastic data generative model.
@@ -614,6 +617,7 @@ def set_h0_params(self,
             self._set_h0_params_recursion(self.h0_root,h0_root)
 
         self.reset_hn_params()
+        return self
 
     def get_h0_params(self):
         """Get the hyperparameters of the prior distribution.
@@ -668,6 +672,7 @@ def set_hn_params(self,
             self._set_hn_params_recursion(self.hn_root,hn_root)
 
         self.calc_pred_dist(np.zeros(self.c_d_max,dtype=int))
+        return self
 
     def get_hn_params(self):
         """Get the hyperparameters of the posterior distribution.
@@ -727,6 +732,7 @@ def update_posterior(self,x):
 
         for i in range(x.shape[0]):
             self._update_posterior_recursion(self.hn_root,x,i)
+        return self
 
     def _map_recursion_add_nodes(self,node:_Node):
         if node.depth == self.c_d_max:  # leaf node
@@ -976,6 +982,7 @@ def calc_pred_dist(self,x):
             self.p_theta_vec[:] = self.hn_beta_vec / self.hn_beta_vec.sum()
         else:
             self.p_theta_vec[:] = self._calc_pred_dist_recursion(self.hn_root,x,i)
+        return self
 
     def make_prediction(self,loss="KL"):
         """Predict a new data point under the given criterion.