snapshot...

modules/deeplearning/cloud_opd_srcnn_abi.py

@@ -730,8 +730,7 @@ def run_evaluate_static(in_file, out_file, ckpt_dir):
     cld_opd = cld_opd[nn.slc_y_m, nn.slc_x_m]
     cld_opd = np.expand_dims(cld_opd, axis=0)
     cld_opd = nn.upsample(cld_opd)
-    # cld_opd = normalize(cld_opd, label_param, mean_std_dct)
-    cld_opd = scale(cld_opd, label_param, mean_std_dct)
+    cld_opd = normalize(cld_opd, label_param, mean_std_dct)
     print('OPD done')
 
     data = np.stack([bt, refl, cld_opd], axis=3)
@@ -739,7 +738,6 @@ def run_evaluate_static(in_file, out_file, ckpt_dir):
     h5f.close()
 
     cld_opd_sres = nn.run_evaluate(data, ckpt_dir)
-    # cld_opd_sres = denormalize(cld_opd_sres, label_param, mean_std_dct)
     cld_opd_sres = descale(cld_opd_sres, label_param, mean_std_dct)
     _, ylen, xlen, _ = cld_opd_sres.shape
     print('OUT: ', ylen, xlen)
@@ -754,8 +752,7 @@ def run_evaluate_static(in_file, out_file, ckpt_dir):
     cld_opd_out[0:(ylen+2*border), 0:(xlen+2*border)] = cld_opd[0, :, :]
 
     refl_out = denormalize(refl_out, 'refl_0_65um_nom', mean_std_dct)
-    # cld_opd_out = denormalize(cld_opd_out, label_param, mean_std_dct)
-    cld_opd_out = descale(cld_opd_out, label_param, mean_std_dct)
+    cld_opd_out = denormalize(cld_opd_out, label_param, mean_std_dct)
 
     if out_file is not None:
         np.save(out_file, (cld_opd_sres_out, refl_out, cld_opd_out, cld_opd_hres))