Custom Forecaster¶

In this example, we use a custom forecast algorithm, the "Olympic Average", implemented in Python. In this forecast, the mean (mu) forecast for any bucket is computed by taking the 26 historical values, discarding the highest and lowest, and taking the mean of the remaining 24 values.

In [1]:

import passengersim as pax
pax.versions()

import passengersim as pax pax.versions()

passengersim 0.18.1
passengersim.core 0.18.1

We'll first run a simulation without the change as a baseline. As this tutorial is meant as a technology demonstration and not for serious statistical analysis, we'll only run one trial so it goes fast, and we won't worry about statistical validity.

In [2]:

cfg = pax.Config.from_yaml([
    "network/08-untrunc-em.yaml",
])

cfg = pax.Config.from_yaml([ "network/08-untrunc-em.yaml", ])

In [3]:

cfg.simulation_controls.num_trials = 1

cfg.simulation_controls.num_trials = 1

In [4]:

sim = pax.Simulation(cfg)
baseline_summary = sim.run()

sim = pax.Simulation(cfg) baseline_summary = sim.run()

Task Completed after 2.17 seconds

In [5]:

baseline_summary.fig_carrier_revenues()

baseline_summary.fig_carrier_revenues()

Out[5]:

Now we swap in our custom forecast algorithm.

In [6]:

from passengersim.extra.olympic_forecast import OlympicForecastStep

from passengersim.extra.olympic_forecast import OlympicForecastStep

In [7]:

extra_config = """
rm_systems:
  rm_test2:
    availability_control: leg
    processes:
      DCP:
      - step_type: untruncation
        algorithm: em
        kind: leg
      - step_type: olympicforecast
      - step_type: emsr
        algorithm: emsrb
airlines:
  - name: AL1
    rm_system: rm_test2
  - name: AL2
    rm_system: rm_test1
"""

cfg = pax.Config.from_yaml([
    "network/08-untrunc-em.yaml",
    extra_config,
])
cfg.simulation_controls.num_trials = 1

extra_config = """ rm_systems: rm_test2: availability_control: leg processes: DCP: - step_type: untruncation algorithm: em kind: leg - step_type: olympicforecast - step_type: emsr algorithm: emsrb airlines: - name: AL1 rm_system: rm_test2 - name: AL2 rm_system: rm_test1 """ cfg = pax.Config.from_yaml([ "network/08-untrunc-em.yaml", extra_config, ]) cfg.simulation_controls.num_trials = 1

In [8]:

sim = pax.Simulation(cfg)

sim = pax.Simulation(cfg)

In [9]:

summary = sim.run()

summary = sim.run()

Task Completed after 13.70 seconds

In [10]:

summary.fig_carrier_revenues()

summary.fig_carrier_revenues()

Out[10]:

Comparing Results¶

In [11]:

from passengersim.contrast import Contrast

comp = Contrast(Baseline=baseline_summary, Olympic=summary)

from passengersim.contrast import Contrast comp = Contrast(Baseline=baseline_summary, Olympic=summary)

In [12]:

comp.fig_carrier_revenues()

comp.fig_carrier_revenues()

Out[12]:

In [13]:

comp.fig_carrier_load_factors()

comp.fig_carrier_load_factors()

Out[13]:

In [14]:

comp.fig_fare_class_mix()

comp.fig_fare_class_mix()

Out[14]:

In [15]:

comp.fig_bookings_by_timeframe(by_carrier="AL1", by_class=True, source_labels=True)

comp.fig_bookings_by_timeframe(by_carrier="AL1", by_class=True, source_labels=True)

Out[15]:

In [16]:

comp.fig_leg_forecasts(by_flt_no=101, of=["mu", "sigma"])

comp.fig_leg_forecasts(by_flt_no=101, of=["mu", "sigma"])

Out[16]:

In [17]:

comp.fig_leg_forecasts(by_flt_no=111, of=["mu", "sigma"])

comp.fig_leg_forecasts(by_flt_no=111, of=["mu", "sigma"])

Out[17]: