Example scenarios

The epifx.example.seir module provides example scenarios for each of the provided Epidemic models, using Google Flu Trends data for Australia in 2014.

Use the write_example_files() function to write the scenario definitions and input data files for a specific scenario to the current working directory. You can then run forecast simulations with the epifx-forecast command. You can also modify the scenario definitions and input data files to see how your changes affect the forecast results.

epifx.example.seir.write_example_files(scenario)

Save the example files for a scenario to the working directory.

Parameters:

scenario – The scenario name.

Raises:

ValueError – If the scenario name is invalid (see below).

The valid scenarios are:

  • 'seir': The deterministic SEIR model with date times.

  • 'seir_quick': : The deterministic SEIR model with date times, using only 20 particles.

  • 'seeiir': The deterministic SEEIIR model with date times.

  • 'seeiir_scalar': The deterministic SEEIIR model with scalar time.

  • 'stoch': The stochastic SEEIIR model with date times.

Shown below is an example script that generates forecasts for the 'seir' scenario:

import datetime
import epifx.example.seir
import pypfilt

scenario = 'seir'
toml_file = 'seir.toml'
output_file = 'seir_forecasts.hdf5'
forecast_dates = [datetime.datetime(2014, 4, 1)]

# Write the scenario files to the working directory.
epifx.example.seir.write_example_files(scenario)

# Run forecasts for each scenario instance (there is only one instance).
for instance in pypfilt.load_instances(toml_file):
    context = instance.build_context()
    pypfilt.forecast(context, forecast_dates, output_file)