Merge pull request #30 from ziatdinovmax/dev

Dev

gpim/gpbayes/boptim.py

@@ -144,6 +144,10 @@ class boptimizer:
         **extent(list of lists):
             Define bounds for multi-dimensional data. For example, for 2D data,
             the extent parameter is [[xmin, xmax], [ymin, ymax]]
+        **save_checkpoints (bool):
+            save results to disk after each iteration
+        **filename (str):
+            name of a file to save checkpoints to
         **verbose (int):
             Level of verbosity (0, 1, or 2)
     """
@@ -171,7 +175,6 @@ def __init__(self,
         learning_rate = kwargs.get("learning_rate", 5e-2)
         jitter = kwargs.get("jitter", 1.0e-6)
         isotropic = kwargs.get("isotropic", False)
-
         self.precision = kwargs.get("precision", "double")
 
         if self.use_gpu and torch.cuda.is_available():
@@ -213,6 +216,8 @@ def __init__(self,
         self.points_mem = kwargs.get("memory", 10)
         self.exit_strategy = kwargs.get("exit_strategy", 0)
         self.mask = kwargs.get("mask", None)
+        self.save_checkpoints = kwargs.get("save_checkpoints", False)
+        self.filename = kwargs.get("filename", "./boptim_results")
         self.indices_all, self.vals_all = [], []
         self.target_func_vals, self.gp_predictions = [y_seed.copy()], []
 
@@ -375,23 +380,23 @@ def dist(idx):
 
         dscale_ = 0 if self.dscale is None else self.dscale
         _idx = 0
-        if self.verbose:
+        if self.verbose == 2:
             print('Acquisition function max value {} at {}'.format(
                 val_list[_idx], idx_list[_idx]))
         if len(self.indices_all) == 0:
             return idx_list[_idx], val_list[_idx]
         while (1 in [1 for a in self.indices_all if a == idx_list[_idx]]
                 or dist(idx_list[_idx])):
-            if self.verbose:
+            if self.verbose == 2:
                 print("Finding the next max point...")
             _idx = _idx + 1
             if _idx == len(idx_list):
                 _idx = np.random.randint(0, len(idx_list)) if self.exit_strategy else -1
-                if self.verbose:
+                if self.verbose == 2:
                     print('Index out of list. Exiting with acquisition function value {} at {}'.format(
                         val_list[_idx], idx_list[_idx]))
                 break
-            if self.verbose:
+            if self.verbose == 2:
                 print('Acquisition function max value {} at {}'.format(
                     val_list[_idx], idx_list[_idx]))
         return idx_list[_idx], val_list[_idx]
@@ -426,6 +431,24 @@ def run(self):
         """
         for i in range(self.exploration_steps):
             self.single_step(i)
+            if self.save_checkpoints:
+                self.save_results()
+        self.save_results()
         if self.verbose:
             print("\nExploration completed")
         return
+
+    def save_results(self, *args):
+        """
+        Save indermediary and/or final results
+        """
+        try:
+            filename = args[0]
+        except IndexError:
+            filename = self.filename
+        results = {}
+        results['gp_pred'] = self.gp_predictions
+        results['func_val'] = self.target_func_vals
+        results['inds_all'] = np.array(self.indices_all)
+        results['vals_all'] = np.array(self.vals_all)
+        np.save(filename+".npy", results)

setup.py

@@ -1,6 +1,6 @@
 __author__ = "Maxim Ziatdinov"
 __copyright__ = "Copyright Maxim Ziatdinov (2020)"
-__version__ = "0.3.4"
+__version__ = "0.3.5"
 __maintainer__ = "Maxim Ziatdinov"
 __email__ = "[email protected]"
 __date__ = "05/20/2020"