Update docstring

flare/gp.py

@@ -122,10 +122,16 @@ def force_list_to_np(forces: list):
         return forces_np
 
     def train(self, output=None, opt_param_override: dict = None):
-        """Train Gaussian Process model on training data. Tunes the \
+        """
+        Train Gaussian Process model on training data. Tunes the \
 hyperparameters to maximize the likelihood, then computes L and alpha \
-(related to the covariance matrix of the training set)."""
+(related to the covariance matrix of the training set).
 
+        :param output: Output to write to
+        :param opt_param_override: Dictionary of parameters to override
+        instace's optimzation parameters.
+        :return:
+        """
         x_0 = self.hyps
 
         args = (self.training_data, self.training_labels_np,
@@ -141,8 +147,8 @@ def train(self, output=None, opt_param_override: dict = None):
 
         grad_tol = opt_param_temp.get('grad_tol', 1e-4)
         x_tol = opt_param_temp.get('x_tol', 1e-5)
-        line_steps = opt_param_temp.get('maxls', 20)
-        max_iter = opt_param_temp.get('maxiter', self.maxiter)
+        line_steps = opt_param_temp.get('max_ls', 20)
+        max_iter = opt_param_temp.get('max_iter', self.maxiter)
         algo = opt_param_temp.get('algorithm', self.algo)
         disp = opt_param_temp.get('disp', False)
 
@@ -164,7 +170,6 @@ def train(self, output=None, opt_param_override: dict = None):
                 print("Warning! Algorithm for L-BFGS-B failed. Changing to "
                       "BFGS for remainder of run.")
                 self.opt_params['algorithm'] = 'BFGS'
-                algo = 'BFGS'
 
         if opt_param_temp.get('custom_bounds', None) is not None:
             res = minimize(get_neg_like_grad, x_0, args,