traveltimes_prediction.interface package¶
Submodules¶
traveltimes_prediction.interface.prediction_manager module¶
-
class
traveltimes_prediction.interface.prediction_manager.
PredictionManager
[source]¶ Bases:
object
Class managing the prediction routine for all allowed sections.
-
launch_prediction
(sections=['TEST-TEST'], prediction_interval_s=60, test_dtime=None)[source]¶ Method for looped prediction of traveltime.
Parameters: - sections (list) – list of all sections for which the prediction should be executed.
- prediction_interval_s (number) – how often should the manager execute prediction.
- test_dtime (datetime) – For TESTING PURPOSES – artificial “now” datetime.
-
traveltimes_prediction.interface.section_interface module¶
-
class
traveltimes_prediction.interface.section_interface.
SectionInterface
(section, models)[source]¶ Bases:
object
Class - interface of the individual sections. Serves to train models for sections and predict using these models.
-
predict
(time_interval, db_interface)[source]¶ Method for prediction of the traveltime, taking data from given time_interval.
Parameters: - time_interval (dict) – {‘from’: datetime, ‘to’: datetime}
- db_interface (object) – instance of DBInterface
Returns: list of dict - info about the process of prediction
-
traveltimes_prediction.interface.training_manager module¶
-
class
traveltimes_prediction.interface.training_manager.
TrainingManager
[source]¶ Bases:
object
Class for training of the models for specified sections.
-
prepare_training_data
(section, time_interval_list)[source]¶ Method for retrieving the data from database and preparing them for training.
Parameters: - section (string) – e.g. ‘KOCE-LNCE’
- time_interval_list (list) – list of dicts - {‘from: datetime, ‘to’: datetime}
Returns: tuple - (X, Y) - data for training
-
train
(section, MODEL, model_params, X_Y_data)[source]¶ Method for training of the model.
Parameters: - section (string) – e.g. ‘LNCE-KOCE’
- MODEL (class) – class of the model to be created - e.g. ClusterModel, TimeDomainModel...
- model_params (dict) – model parameters
- X_Y_data (tuple) – tuple of DataFrames - input data for training
-
training_loop
(sections=['TEST-TEST'], training_interval_s=86400, train_for=None)[source]¶ Method for the training to be kept in loop, calls self.train().
Parameters: - sections (list) – list of section whose models are gonna be trained.
- training_interval_s (int) – hours, how often should be the models trained.
- train_for (datetime.datetime) – for TEST purposes, datetime for which should be trained.
-