Adaptions and corrections to config_postprocess.py.

video_prediction_tools/config_runscripts/config_postprocess.py

@@ -30,8 +30,8 @@ class Config_Postprocess(Config_runscript_base):
         self.checkpoint_dir = None
         self.destination_dir = None
         # list of variables to be written to runscript
-        self.list_batch_vars = ["VIRT_ENV_NAME", "source_dir", "destination_dir",
-                                "checkpoint_dir", "model", "dataset"]
+        self.list_batch_vars = ["VIRT_ENV_NAME", "source_dir", "results_dir",
+                                "checkpoint_dir", "model"]
         # copy over method for keyboard interaction
         self.run_config = Config_Postprocess.run_postprocess
     #
@@ -43,7 +43,7 @@ class Config_Postprocess(Config_runscript_base):
         :return: all attributes of class postprocess are set
         """
         # decide which dataset is used
-        dset_type_req_str = "Enter the name of the dataset on which training was performed:\n"
+        dset_type_req_str = "Enter the name of the dataset on which training was performed:"
         dset_err = ValueError("Please select a dataset from the ones listed above.")
 
         self.dataset = Config_Postprocess.keyboard_interaction(dset_type_req_str, Config_Postprocess.check_dataset,
@@ -56,21 +56,22 @@ class Config_Postprocess(Config_runscript_base):
         # get the 'checkpoint-directory', i.e. the directory where the trained model parameters are stored
         # Note that the remaining information (model, results-directory etc.) can be retrieved form it!!!
         trained_dir_req_str = "Enter the absolute (!) path to the model checkpoint directory" + \
-                              " for which postprocessing should be done:\n"
+                              " for which postprocessing should be done:"
         trained_err = FileNotFoundError("No trained model parameters found.")
 
         self.checkpoint_dir = Config_Postprocess.keyboard_interaction(trained_dir_req_str,
                                                                       Config_Postprocess.check_traindir,
                                                                       trained_err, ntries=3)
 
-        # get the relevant information from checlpoint_dir in order to construct source_dir and results_dir
+        # get the relevant information from checkpoint_dir in order to construct source_dir and results_dir
         # (following naming convention)
         cp_dir_split = Config_Postprocess.path_rec_split(self.checkpoint_dir)
+        cp_dir_split = list(filter(None, cp_dir_split))                       # get rid of empty list elements
 
-        base_dir, exp_dir_base, exp_dir = os.path.join(*cp_dir_split[:-3]), cp_dir_split[-3], cp_dir_split[-1]
+        base_dir, exp_dir_base, exp_dir = "/"+os.path.join(*cp_dir_split[:-4]), cp_dir_split[-3], cp_dir_split[-1]
         self.model = Config_Postprocess.check_model(cp_dir_split[-2])
 
-        self.source_dir = Config_Postprocess.check_source(os.path.join(base_dir, exp_dir_base))
+        self.source_dir = Config_Postprocess.check_source(os.path.join(base_dir, "preprocessedData", exp_dir_base))
         self.destination_dir = os.path.join(base_dir, "results", exp_dir_base, self.model, exp_dir)
     #
     # -----------------------------------------------------------------------------------
@@ -128,7 +129,7 @@ class Config_Postprocess(Config_runscript_base):
         """
         if not model_in in Config_Postprocess.list_models:
             print("**** Known models ****")
-            for model in Config_Postprocess: print(model)
+            for model in Config_Postprocess.list_models: print(model)
             raise ValueError("{0} is an unknown model (see list of known models above).".format(model_in))
         else:
             pass
@@ -144,7 +145,7 @@ class Config_Postprocess(Config_runscript_base):
         :param source_dir_in: input directory to be checked
         :return: returns source_dir_in when check is passed successfully
         """
-        real_dir = os.path.join(source_dir_in, "tfrecords")
+        real_dir = os.path.join(source_dir_in, "tfrecords", "tfrecords")
         if os.path.isdir(real_dir):
             file_list = glob.glob(os.path.join(real_dir, "sequence*.tfrecords"))
             if len(file_list) > 0: