Forecasts

leg_forecasts

This table shows the demand forecasts made for each booking class on each leg as departure approaches. Each row identifies a carrier, leg, booking class, and data collection point (DCP). The values are averages across samples, excluding burned samples.

Alongside the forecast mean and standard deviation, the table includes measures of how often inventory was closed in the history used to make the forecast. These can help you understand the information the forecasting process had available. The forecast standard deviation column is an average of the reported standard deviations, rather than a new standard deviation calculated from all the forecasts together.

If it is available, you can access this table as summary.leg_forecasts.

Index

Field

Type

Meaning

carrier

string

Carrier code.

leg_id

integer

Leg identifier; joins to legs.

bucket_number

integer

Leg bucket number.

booking_class

string

Booking class code.

days_prior

integer

Days before departure at the recorded observation or data collection point (DCP).

Columns

Field

Type

Meaning

forecast_mean

number

Mean of the recorded demand forecast means, in passengers.

forecast_stdev

number

Mean of the recorded forecast standard deviations; not a pooled standard deviation.

forecast_closed_in_tf

number

Mean closure fraction for this timeframe in the history used for forecasting.

forecast_closed_in_future

number

Mean closure fraction for future timeframes in the history used for forecasting.

The values reflect the forecasts produced by the carrier’s configured revenue-management system. Recording the table does not create forecasts that the system itself does not calculate.

Activation

This table is built from records saved in the simulation database. To make it available, set db.engine to sqlite and include bucket in db.write_items before running the simulation. Add leg_forecasts to outputs.reports if you want PassengerSim to prepare the table as part of the run.

For example, your configuration could include the following. These lists replace the existing settings, including the default write_items, so keep any other entries you still need:

db:
  engine: sqlite
  write_items: [bucket]
outputs:
  reports: [leg_forecasts]

Use bucket to record forecasts throughout the booking horizon. The bucket_final option only provides records at departure.

To stop preparing this table automatically, remove leg_forecasts from outputs.reports. You can still access it later if the required database records remain available. To stop recording the supporting data as well, remove bucket from db.write_items, along with any other recording options that save the same data. Keep those options if another report still needs them.

See database-backed tables for how to keep the supporting database available or save a prepared table with your summary.

path_forecasts

This table shows demand forecasts at the itinerary and booking-class level. Each row represents a booking class on a path at a particular data collection point (DCP), with values averaged across samples excluding burned samples.

In addition to the forecast and closure measures, it includes the average adjusted price recorded for the path class. The values depend on the revenue-management processes you configured. As with leg_forecasts, forecast_stdev is an average of the reported standard deviations, not a pooled measure of variation.

If it is available, you can access this table as summary.path_forecasts.

Index

Field

Type

Meaning

path_id

integer

Path identifier; joins to paths.

booking_class

string

Booking class code.

days_prior

integer

Days before departure at the recorded observation or data collection point (DCP).

Columns

Field

Type

Meaning

forecast_mean

number

Mean of the recorded demand forecast means, in passengers.

forecast_stdev

number

Mean of the recorded forecast standard deviations; not a pooled standard deviation.

forecast_closed_in_tf

number

Mean closure fraction for this timeframe in the history used for forecasting.

forecast_closed_in_future

number

Mean closure fraction for future timeframes in the history used for forecasting.

adjusted_price

number

Mean recorded adjusted price, in simulation currency.

Recording these results does not enable a forecasting or pricing method by itself. The values come from the revenue-management processes already configured for the path.

Activation

This table is built from records saved in the simulation database. To make it available, set db.engine to sqlite and include pathclass in db.write_items before running the simulation. Add path_forecasts to outputs.reports if you want PassengerSim to prepare the table as part of the run.

For example, your configuration could include the following. These lists replace the existing settings, including the default write_items, so keep any other entries you still need:

db:
  engine: sqlite
  write_items: [pathclass]
outputs:
  reports: [path_forecasts]

Use pathclass to record the forecasts at each DCP. The pathclass_final option only keeps departure observations.

To stop preparing this table automatically, remove path_forecasts from outputs.reports. You can still access it later if the required database records remain available. To stop recording the supporting data as well, remove pathclass from db.write_items, along with any other recording options that save the same data. Keep those options if another report still needs them.

See database-backed tables for how to keep the supporting database available or save a prepared table with your summary.