snapshot...

modules/deeplearning/espcn.py

@@ -49,16 +49,17 @@ f.close()
 mean_std_dct.update(mean_std_dct_l1b)
 mean_std_dct.update(mean_std_dct_l2)
 
-emis_params = ['temp_10_4um_nom', 'temp_11_0um_nom', 'temp_12_0um_nom', 'temp_13_3um_nom', 'temp_3_75um_nom',
-               'temp_6_7um_nom', 'temp_6_2um_nom', 'temp_7_3um_nom', 'temp_8_5um_nom', 'temp_9_7um_nom']
-l2_params = ['cloud_fraction', 'cld_temp_acha', 'cld_press_acha', 'cld_opd_acha', 'cld_reff_acha']
+# emis_params = ['temp_10_4um_nom', 'temp_11_0um_nom', 'temp_12_0um_nom', 'temp_13_3um_nom', 'temp_3_75um_nom',
+#                'temp_6_7um_nom', 'temp_6_2um_nom', 'temp_7_3um_nom', 'temp_8_5um_nom', 'temp_9_7um_nom']
+data_params = ['refl_0_65um_nom', 'temp_11_0um_nom', 'cld_temp_acha', 'cld_press_acha', 'cloud_fraction']
+label_params = ['refl_0_65um_nom', 'temp_11_0um_nom', 'cld_temp_acha', 'cld_press_acha', 'cloud_fraction']
+
 
-# -- Zero out params (Experimentation Only) ------------
-zero_out_params = ['cld_reff_dcomp', 'cld_opd_dcomp', 'iwc_dcomp', 'lwc_dcomp']
 DO_ZERO_OUT = False
 
-label_idx = 1
-label_param = l2_params[label_idx]
+data_idx, label_idx = 1, 1
+data_param = data_params[data_idx]
+label_param = label_params[label_idx]
 
 
 def build_conv2d_block(conv, num_filters, activation, block_name, padding='SAME'):
@@ -211,12 +212,12 @@ class ESPCN:
 
         self.n_chans = 1
 
-        # self.X_img = tf.keras.Input(shape=(None, None, self.n_chans))
-        self.X_img = tf.keras.Input(shape=(36, 36, self.n_chans))
+        self.X_img = tf.keras.Input(shape=(None, None, self.n_chans))
+        # self.X_img = tf.keras.Input(shape=(36, 36, self.n_chans))
 
         self.inputs.append(self.X_img)
-        # self.inputs.append(tf.keras.Input(shape=(None, None, self.n_chans)))
-        self.inputs.append(tf.keras.Input(shape=(36, 36, self.n_chans)))
+        self.inputs.append(tf.keras.Input(shape=(None, None, self.n_chans)))
+        # self.inputs.append(tf.keras.Input(shape=(36, 36, self.n_chans)))
 
         self.DISK_CACHE = False
 
@@ -247,17 +248,24 @@ class ESPCN:
         label = label[:, label_idx, :, :]
         label = np.expand_dims(label, axis=3)
 
+        data = data[:, data_idx, :, :]
+        data = np.expand_dims(data, axis=3)
+
         data = data.astype(np.float32)
         label = label.astype(np.float32)
 
-        data_norm = []
-        for k, param in enumerate(emis_params):
-            tmp = normalize(data[:, k, :, :], param, mean_std_dct)
-            data_norm.append(tmp)
-        data = np.stack(data_norm, axis=3)
+        # data_norm = []
+        # for k, param in enumerate(emis_params):
+        #     tmp = normalize(data[:, k, :, :], param, mean_std_dct)
+        #     data_norm.append(tmp)
+        # data = np.stack(data_norm, axis=3)
+        #
+        # if label_param != 'cloud_fraction':
+        #     label = scale(label, label_param, mean_std_dct)
 
+        data = normalize(data, data_param, mean_std_dct)
         if label_param != 'cloud_fraction':
-            label = scale(label, label_param, mean_std_dct)
+            label = normalize(label, label_param, mean_std_dct)
 
         if is_training and DO_AUGMENT:
             data_ud = np.flip(data, axis=1)