Getting Started#

Power Balance Models (PBM) can be run via a command line interface (CLI) or by constructing scripts to access the API. Whichever case is chosen runs can be customised by the user as much (or as little) as required.

Example run#

As there are existing configurations within the API you can run PBM 'out the box' using the available CLI command:

powerbalance run

after the run is completed a new timestamped directory will be produced containing not only the results but a preserved copy of the configuration set to achieve them. This is useful for recreating runs.

In addition your chosen browser should be launched to show a webpage displaying the power data plots as dynamic widgets which can be interacted with.

Viewing existing plots#

A new plot browser window can be opened on any results directory using the view-results command:

powerbalance view-results <pbm_results_dir>

Viewing the data#

Data is written to a HDF5 file, with each model's output being written to the file as a single dataframe the key for which is the model name in lower case with any . being replaced with _.

The recommended method for accessing the data is using the Pandas module. In the case of the Tokamak.Interdependencies model:

import pandas as pd
import glob
import os
import argparse

# Create a parser so we can select the result directory
# we which to view from the command line
parser = argparse.ArgumentParser()
parser.add_argument('result_dir')
args = parser.parse_args()

# Use glob to lazily find the HDF5 file
hdf5_file = glob.glob(
    os.path.join(args.result_dir, 'data', '*.hdf5')
)[0]

# Accessing the dataframe using the key within the file
data_frame = pd.read_hdf(hdf5_file, key='tokamak_interdependencies')

print(data_frame)

Creating your own scripts#

The main class used to initialise and run a simulation via the OpenModelica backend within PBM is the PowerBalance class. It is recommended that the class be used via a context manager to ensure that model build directories are removed upon completion (this is done automatically when using the CLI).

from power_balance.core import PowerBalance

# Initialise the PBM class fetching parameters
# and models from the default locations
# Use context manager to ensure junk collection
with PowerBalance() as pbm_instance:

    # Run the simulation with the configuration
    pbm_instance.run_simulation()

    # Open the plots in the browser window
    pbm_instance.launch_browser()