snapshot...

modules/deeplearning/cloud_fraction_fcn_abi.py

@@ -777,6 +777,7 @@ def run_evaluate_static(in_file, out_file, ckpt_dir):
     h5f = h5py.File(in_file, 'r')
 
     bt = get_grid_values_all(h5f, 'temp_11_0um_nom')
+    y_len, x_len = bt.shape
     refl = get_grid_values_all(h5f, 'refl_0_65um_nom')
     refl_lo = get_grid_values_all(h5f, 'refl_0_65um_nom_min_sub')
     refl_hi = get_grid_values_all(h5f, 'refl_0_65um_nom_max_sub')
@@ -785,43 +786,35 @@ def run_evaluate_static(in_file, out_file, ckpt_dir):
     lons = get_grid_values_all(h5f, 'longitude')
     lats = get_grid_values_all(h5f, 'latitude')
 
-    # bt = bt[0:2500, :]
-    # refl = refl[0:2500, :]
-    # lons = lons[0:2500, :]
-    # lats = lats[0:2500, :]
-    # refl_lo = refl_lo[0:2500, :]
-    # refl_hi = refl_hi[0:2500, :]
-    # refl_std = refl_std[0:2500, :]
-    # cp = cp[0:2500, :]
+    bt_nh = bt[0:2501, :]
+    refl_nh = refl[0:2501, :]
+    refl_lo_nh = refl_lo[0:2501, :]
+    refl_hi_nh = refl_hi[0:2501, :]
+    refl_std_nh = refl_std[0:2501, :]
+    cp_nh = cp[0:2501, :]
 
-    y_len, x_len = bt.shape[0], bt.shape[1]
-
-    bt = normalize(bt, 'temp_11_0um_nom', mean_std_dct)
-    refl = normalize(refl, 'refl_0_65um_nom', mean_std_dct)
-    refl_lo = normalize(refl_lo, 'refl_0_65um_nom', mean_std_dct)
-    refl_hi = normalize(refl_hi, 'refl_0_65um_nom', mean_std_dct)
-    refl_std = np.where(np.isnan(refl_std), 0, refl_std)
-    cp = np.where(np.isnan(cp), 0, cp)
-
-    data = np.stack([bt, refl, refl_lo, refl_hi, refl_std, cp], axis=2)
-    data = np.expand_dims(data, axis=0)
+    bt_sh = bt[2499:y_len, :]
+    refl_sh = refl[2499:y_len, :]
+    refl_lo_sh = refl_lo[2499:y_len, :]
+    refl_hi_sh = refl_hi[2499:y_len, :]
+    refl_std_sh = refl_std[2499:y_len, :]
+    cp_sh = cp[2499:y_len, :]
 
     h5f.close()
 
-    nn = SRCNN()
-    probs = nn.run_evaluate(data, ckpt_dir)
-    cld_frac = probs.argmax(axis=3)
-    cld_frac = cld_frac.astype(np.int8)
+    cld_frac_nh = run_evaluate_static_(bt_nh, refl_nh, refl_lo_nh, refl_hi_nh, refl_std_nh, cp_nh, ckpt_dir)
+    cld_frac_sh = run_evaluate_static_(bt_sh, refl_sh, refl_lo_sh, refl_hi_sh, refl_std_sh, cp_sh, ckpt_dir)
+
     cld_frac_out = np.zeros((y_len, x_len), dtype=np.int8)
     border = int((KERNEL_SIZE - 1)/2)
-    cld_frac_out[border:y_len - border, border:x_len - border] = cld_frac[0, :, :]
+    cld_frac_out[border:2500, border:x_len - border] = cld_frac_nh[0, :, :]
+    cld_frac_out[2500:y_len - border, border:x_len - border] = cld_frac_sh[0, :, :]
 
     bt = denormalize(bt, 'temp_11_0um_nom', mean_std_dct)
     refl = denormalize(refl, 'refl_0_65um_nom', mean_std_dct)
 
     var_names = ['cloud_fraction', 'temp_11_0um', 'refl_0_65um']
     dims = ['num_params', 'y', 'x']
-
     da = xr.DataArray(np.stack([cld_frac_out, bt, refl], axis=0), dims=dims)
     da.assign_coords({
         'num_params': var_names,
@@ -835,6 +828,24 @@ def run_evaluate_static(in_file, out_file, ckpt_dir):
         return [cld_frac_out, bt, refl, cp, lons, lats]
 
 
+def run_evaluate_static_(bt, refl, refl_lo, refl_hi, refl_std, cp, ckpt_dir):
+    bt = normalize(bt, 'temp_11_0um_nom', mean_std_dct)
+    refl = normalize(refl, 'refl_0_65um_nom', mean_std_dct)
+    refl_lo = normalize(refl_lo, 'refl_0_65um_nom', mean_std_dct)
+    refl_hi = normalize(refl_hi, 'refl_0_65um_nom', mean_std_dct)
+    refl_std = np.where(np.isnan(refl_std), 0, refl_std)
+    cp = np.where(np.isnan(cp), 0, cp)
+
+    data = np.stack([bt, refl, refl_lo, refl_hi, refl_std, cp], axis=2)
+    data = np.expand_dims(data, axis=0)
+    nn = SRCNN()
+    probs = nn.run_evaluate(data, ckpt_dir)
+    cld_frac = probs.argmax(axis=3)
+    cld_frac = cld_frac.astype(np.int8)
+
+    return cld_frac
+
+
 def analyze_3cat(file):
 
     tup = np.load(file, allow_pickle=True)