Make time series possible for other attributes

run_pypsa.py

@@ -114,31 +114,31 @@ def dicts_to_pypsa(case_dict, component_list, component_attr):
     n = add_buses_to_network(n, component_list)
 
     for component_dict in component_list:
-        for attribute in component_dict:
-            # Check if attribute is a string and csv file holding a time series
-            if isinstance(component_dict[attribute], str) and component_dict[attribute].endswith('.csv'):
-                logging.info(f"Processing time series file: {component_dict[attribute]}")
-                # Add time series to components
-                ts_file = os.path.join(case_dict["input_path"],component_dict[attribute])
+        # for generators and loads, add time series to components
+        for attr in component_dict:
+            # Add time series to components
+            if isinstance(component_dict[attr], str) and ".csv" in component_dict[attr]:
+                logging.info("reading time series file")
+                ts_file = os.path.join(case_dict["input_path"],component_dict[attr])
                 try:
                     ts = process_time_series_file(ts_file, case_dict["datetime_start"], case_dict["datetime_end"])
-                    logging.info(f"Time series file: {component_dict[attribute]} processed successfully.")
-                    logging.info(ts)
                 except Exception:  # if time series not found in input path, use csv's in test directory
-                    logging.error("Time series file not found for " + component_dict[attribute] + " of " + component_dict["name"] + ". Now exiting.")
-                    sys.exit(1)
+                    logging.warning("Time series file not found for " + component_dict["name"] + ". Using time series files in test directory.")
+                    case_dict['input_path'] = "./test"
+                    ts_file = os.path.join(case_dict["input_path"],component_dict[attr])
+                    ts = process_time_series_file(ts_file, case_dict["datetime_start"], case_dict["datetime_end"])
 
                 if ts is not None:
                     # Include time series as snapshots taking every delta_t value
                     n.snapshots = ts.iloc[::case_dict['delta_t'], :].index if case_dict['delta_t'] else ts.index
                     # Add time series to component
-                    component_dict[attribute] = ts.iloc[:, 0]
+                    component_dict[attr] = ts.iloc[:, 0]
+
                     # Scale by numerics_scaling, this avoids rounding otherwise done in Gurobi for small numbers and normalize time series if needed
-                    component_dict = scale_normalize_time_series(component_dict, case_dict["numerics_scaling"])
-                    # Remove time_series_file from component_dict
+                    component_dict = scale_normalize_time_series(component_dict, case_dict["numerics_scaling"])                 
                 else:
-                    logging.warning("Time series not properly processed for " + component_dict[attribute] + " of " + component_dict["name"] + ". Now exiting.")
-                    sys.exit(1)
+                    logging.warning("Time series file not found for " + component_dict["name"] + ". Skipping component.")
+                    continue
 
         # Without time series file, set snaphsots to number of time steps defined in the input file
         if len(n.snapshots) == 1 and case_dict["no_time_steps"] is not None:
@@ -230,7 +230,7 @@ def postprocess_results(n, case_dict):
     time_results_df = pd.concat([time_results_df, n.links_t["p0"].rename(columns=dict(
         zip(n.links_t["p0"].columns.to_list(),
             [name + " dispatch" for name in n.links_t["p0"].columns.to_list()])))], axis=1)
-    
+
     # Collect objective and system cost in one dataframe
     system_cost = (n.statistics()["Capital Expenditure"].sum() + n.statistics()[
         "Operational Expenditure"].sum()) / case_dict["total_hours"]

utilities/read_input.py

@@ -130,8 +130,8 @@ def read_component_data(comp_dict, attr, val, technology, costs_df):
     # if it's empty or a cost name, use read_attr to get the value from the costs dataframe.
     if attr != None:
         read_attr = None
-        # if "name", "bus", or "time_series_file" is in attr or value can be converted to a float, use that
-        if (val != None and (any(x in attr for x in ['name', 'bus', 'carrier']) or is_number(val) or '=' in val)):
+        # if "name", "bus" or "carrier" is in attr or value can be converted to a float, use that
+        if (val != None and (any(x in attr for x in ['name', 'bus', 'carrier']) or is_number(val) or '=' in val or '.csv' in val)):
             comp_dict[attr] = val
         # if otherwise value is a string, use database value if the string is just 'db'
         # if first two letters are db use the rest of the string as the attribute name
@@ -147,13 +147,12 @@ def read_component_data(comp_dict, attr, val, technology, costs_df):
                 else:
                     val = val.replace('db_','')
                     read_attr = val
-            elif val.endswith('.csv'):
-                comp_dict[attr] = val
             else:
-                logging.error('Failed to read in '+val + ' for attribute ' + attr + ' for component ' + comp_dict["component"] + ' ' + comp_dict["name"])
-                logging.error('Exiting now.')
+                logging.error('Tried to read in a string that is not a number, name, or contains "db" to indicate use a database value. Failed = '+val + ' for attribute ' + attr + ' for component ' + comp_dict["component"] + ' ' + comp_dict["name"])
+                logging.error('Terminal error. Exiting.')
                 exit()
 
+
         # if read_attr is defined, use it to get the value from the costs dataframe
         if read_attr != None:
             if (technology in costs_df.index and read_attr in costs_df.columns):