snapshot...

modules/deeplearning/icing_fcn.py

@@ -239,10 +239,11 @@ class IcingIntensityFCN:
         self.test_probs = None
 
         self.learningRateSchedule = None
-        self.num_data_samples = None
+        self.num_data_samples = 1
         self.initial_learning_rate = None
 
         self.data_dct = None
+        self.cth_max = None
 
         n_chans = len(self.train_params)
         if TRIPLET:
@@ -446,7 +447,11 @@ class IcingIntensityFCN:
 
         # TODO: altitude data will be specified by user at run-time
         nda = np.zeros([nd_idxs.size])
+        if self.cth_max is not None:
+            nda[:] = self.cth_max[nd_idxs]
+        else:
             nda[:] = self.flight_level
+
         nda = tf.one_hot(nda, 5).numpy()
         nda = np.expand_dims(nda, axis=1)
         nda = np.expand_dims(nda, axis=1)
@@ -587,6 +592,7 @@ class IcingIntensityFCN:
 
     def setup_eval_pipeline(self, data_dct, num_tiles=1):
         self.data_dct = data_dct
+        self.cth_max = data_dct.get('cth_high_avg', None)
         idxs = np.arange(num_tiles)
         self.num_data_samples = idxs.shape[0]
 
@@ -1058,7 +1064,7 @@ class IcingIntensityFCN:
             pred_s.append(pred)
 
         preds = np.concatenate(pred_s)
-        preds = preds[:,0]
+        preds = np.squeeze(preds)
         self.test_probs = preds
 
         if NumClasses == 2:
@@ -1213,8 +1219,7 @@ def run_evaluate_static(data_dct, num_tiles, ckpt_dir_s_path, day_night='DAY', l
     return preds_dct, probs_dct
 
 
-def run_evaluate_static_2(model, data_dct, num_tiles, day_night='DAY', l1b_or_l2='both', satellite='GOES16',
-                          prob_thresh=0.5, flight_levels=[0, 1, 2, 3, 4], use_flight_altitude=False):
+def run_evaluate_static_2(model, data_dct, num_tiles, prob_thresh=0.5, flight_levels=[0, 1, 2, 3, 4], use_flight_altitude=False):
 
     if not use_flight_altitude:
         flight_levels = [0]
@@ -1233,6 +1238,22 @@ def run_evaluate_static_2(model, data_dct, num_tiles, day_night='DAY', l1b_or_l2
     return preds_dct, probs_dct
 
 
+def load_model(model_path, day_night='NIGHT', l1b_andor_l2='BOTH', satellite='GOES16', use_flight_altitude=False):
+    ckpt_dir_s = os.listdir(model_path)
+    ckpt_dir = model_path + ckpt_dir_s[0]
+
+    model = IcingIntensityFCN(day_night=day_night, l1b_or_l2=l1b_andor_l2, satellite=satellite, use_flight_altitude=use_flight_altitude)
+    model.build_model()
+    model.build_training()
+    model.build_evaluation()
+
+    ckpt = tf.train.Checkpoint(step=tf.Variable(1), model=model.model)
+    ckpt_manager = tf.train.CheckpointManager(ckpt, ckpt_dir, max_to_keep=3)
+    ckpt.restore(ckpt_manager.latest_checkpoint)
+
+    return model
+
+
 if __name__ == "__main__":
     nn = IcingIntensityFCN()
     nn.run('matchup_filename')