initial commit

bo/random.py

@@ -0,0 +1,56 @@
+import numpy as np
+import copy
+
+
+def initialize(config):
+    bo_params = []
+
+    for key, value in config['bo']['params'].items():
+        if value is not None:
+            bo_params.append(key)
+
+    train_x = np.empty((0, len(bo_params)))
+    train_y = np.empty((0,))
+
+    bo_state = {
+        'params': bo_params,
+        'train_x': train_x,
+        'train_y': train_y,
+    }
+
+    return bo_state
+
+
+def ask(config, bo_state):
+    next_config = copy.deepcopy(config)
+
+    for param in bo_state['params']:
+        max_modulation = config['bo']['params'][param]
+        center_value = config['ptycho']['params'][param]
+
+        modulation = max_modulation * (np.random.rand() - 0.5) * 2
+        next_config['ptycho']['params'][param] = center_value + modulation
+
+    return next_config
+
+
+def tell(job_config, bo_state, y_value):
+    x_value = []
+
+    for param in bo_state['params']:
+        x_value.append(job_config['ptycho']['params'][param])
+
+    x_value = np.array(x_value).reshape(1, -1)
+    y_value = np.array([y_value])
+
+    bo_state['train_x'] = np.vstack([
+        bo_state['train_x'],
+        x_value,
+    ])
+
+    bo_state['train_y'] = np.concatenate([
+        bo_state['train_y'],
+        y_value,
+    ])
+
+    return bo_state