search

rl-institut / multi-vector-simulator

Multi-vector Simulation Tool assessing and optimizing Local Energy Systems (LES) for the E-LAND project
GNU General Public License v2.0
21 stars 10 forks source link

Changes in B0_data_input_json.py cause a simulation termination #813

Open TheOneAndra opened 3 years ago

TheOneAndra commented 3 years ago

Changes in B0_data_input_json.py line 29:

    KPI,
    KPI_COST_MATRIX,
    KPI_SCALAR_MATRIX,
    LABEL,

line 343:

 for kpi in (KPI_COST_MATRIX, KPI_SCALAR_MATRIX):

        if not isinstance(dict_values[KPI][kpi], pd.DataFrame):
            dict_values[KPI][kpi] = pd.DataFrame.from_records(dict_values[KPI][kpi])

        if LABEL not in dict_values[KPI][kpi].columns:
            dict_values[KPI][kpi][LABEL] = dict_values[KPI][kpi].index

These changes result in a simulation termination:

(MVS) C:\Andra\MVS_GitHub>python mvs_tool.py -i C:\Andra\RLI\Master_Thesis\Data\MVS\Input_templates\Greenfield_sector_coupling -ext csv -o C:\Andra\RLI\Master_Thesis\Data\MVS\Simulations\Greenfield\210205_sector_coupled_syst\random_test -f -pdf
12:13:28-INFO-Path for logging: C:\Andra\RLI\Master_Thesis\Data\MVS\Simulations\Greenfield\210205_sector_coupled_syst\random_test\mvs_logfile.log
12:13:28-INFO-

 Multi-Vector Simulation Tool (MVS) V0.5.5dev
 Version: 2020-12-18
 Part of the toolbox of H2020 project "E-LAND", Integrated multi-vector management system for Energy isLANDs
 Coded at: Reiner Lemoine Institute (Berlin)
 Contributors: Martha M. Hoffmann

12:13:28-INFO-loading and converting all csv's from C:\Andra\RLI\Master_Thesis\Data\MVS\Input_templates\Greenfield_sector_coupling\csv_elements into one json
c:\andra\mvs_github\src\multi_vector_simulator\A1_csv_to_json.py:283: WrongParameterWarning:

The parameter minimal_degree_of_autonomy in the fileC:\Andra\RLI\Master_Thesis\Data\MVS\Input_templates\Greenfield_sector_coupling\csv_elements\constraints.csv is not expected.
Expected parameters are ['minimal_renewable_factor', 'maximum_emissions']

12:13:28-INFO-Json file created successfully from csv's and stored into C:\Andra\RLI\Master_Thesis\Data\MVS\Input_templates\Greenfield_sector_coupling\csv_elements\mvs_csv_config.json

Traceback (most recent call last):
  File "mvs_tool.py", line 4, in <module>
    main()
  File "c:\andra\mvs_github\src\multi_vector_simulator\cli.py", line 149, in main
    set_default_values=True,
  File "c:\andra\mvs_github\src\multi_vector_simulator\B0_data_input_json.py", line 345, in load_json
    if not isinstance(dict_values[KPI][kpi], pd.DataFrame):
KeyError: 'kpi'

The log stops at: 

2021-03-02 12:13:28,949 - INFO - A1_csv_to_json - Json file created successfully from csv's and stored into C:\Andra\RLI\Master_Thesis\Data\MVS\Input_templates\Greenfield_sector_coupling\csv_elements\mvs_csv_config.json

2021-03-02 12:13:28,955 - DEBUG - A1_csv_to_json - Json created successfully from csv.
2021-03-02 12:13:28,955 - DEBUG - cli - Accessing script: B0_data_input_json

And indeed in _mvs_csvconfig.json there is no KPI.

Would the solution be to add an exception when loading _mvs_csvconfig.json ?

Bachibouzouk commented 3 years ago

A solution would be to ignore if the key KPI is not in the dict_values

--> add an if statement before if not isinstance(dict_values[KPI][kpi], pd.DataFrame):

if KPI in dict_values:
    if not isinstance(dict_values[KPI][kpi], pd.DataFrame):
    # of course indent accordingly