dm.models.predictor package¶
Submodules¶
dm.models.predictor.THPredictorUtil module¶
Support for ventilation length prediction.
Creates training and testing sets for ventilation length prediction, including calculation of distance between a data point and cluster trendline (cluster centroid).
-
dm.models.predictor.THPredictorUtil.
training_testing_data
(data, splitting)¶ It creates training and testing data set.
- Parameters
data – dictionary of data
splitting – the percentage part of data used for training purposes
- Returns
training, testing data sets and minimal number of events that lasted certain time
-
dm.models.predictor.THPredictorUtil.
training_testing_data_only_distance
(training, testing, strategy, strategyFlag, one_line, test_points, cluster_boundaries, cluster_boundaries_all, training_file, testing_file)¶ It creates training and testing data sets including only attributes related to distance.
- Parameters
training – list of training data
testing – list of testing data
strategy – strategy how to compute a slope of line(s)
strategyFlag – flag that denotes a used strategy for computation of a slope of line(s)
one_line – if only one line should be plotted
test_points – true if test points are plotted
cluster_boundaries – if cluster boundaries should be plotted
cluster_boundaries_all – if all cluster boundaries should be plotted
training_file – filename to which training data set is written
testing_file – filename to which training data set is written
- Returns
training and testing data sets including only attributes related to distance
-
dm.models.predictor.THPredictorUtil.
training_testing_data_with_distance
(training, testing, strategy, strategyFlag, one_line, test_points, cluster_boundaries, cluster_boundaries_all, training_file, testing_file)¶ It creates training and testing data sets including attributes related to distance.
- Parameters
training – list of training data
testing – list of testing data
strategy – strategy how to compute a slope of line(s)
strategyFlag – flag that denotes a used strategy for computation of a slope of line(s)
one_line – if only one line should be plotted
test_points – true if test points are plotted
cluster_boundaries – if cluster boundaries should be plotted
cluster_boundaries_all – if all cluster boundaries should be plotted
training_file – filename to which training data set is written
testing_file – filename to which training data set is written
- Returns
training and testing data sets including attributes related to distance
-
dm.models.predictor.THPredictorUtil.
training_testing_data_without_distance
(training, testing, strategy, strategyFlag, one_line, test_points, cluster_boundaries, cluster_boundaries_all, training_file, testing_file)¶ It creates training and testing data sets without attributes related to distance.
- Parameters
training – list of training data
testing – list of testing data
strategy – deprecated
strategyFlag – deprecated
one_line – deprecated
test_points – deprecated
cluster_boundaries – deprecated
cluster_boundaries_all – deprecated
training_file – filename to which training data set is written
testing_file – filename to which training data set is written
- Returns
training and testing data sets without attributes related to distance
dm.models.predictor.generic_training_file module¶
Creates adapted models for predictor of optimal ventilation length.
-
dm.models.predictor.generic_training_file.
create_bins
(data, bins)¶ It creates required number of bins.
- Parameters
data – dictionary of data
bins – number of bins
- Returns
dictionary of data
-
dm.models.predictor.generic_training_file.
training_data
(json_f, cls, devs, model_type, interval_extension, mapper, func, lat, lon, weather)¶ It creates training data.
- Parameters
json_f – dictionary of data
cls – list of clients
devs – list of devices
model_type – type of model - based on CO2 concentration or temperature and humidity
interval_extension – time shift that is subtracted or added to start or end of an interval respectively
mapper – dictionary containing mapping of attribute name in database to the name used in dataset
func – object that determines function used for attribute calculation
lat – latitude of a locality
lon – longitude of a locality
weather – object used to get information about weather
- Returns
dictionary of training data
-
dm.models.predictor.generic_training_file.
training_file_co2
(events_file: str, no_event_time_shift: int, cls, devs, lat, lon, weather)¶ It creates training data for predictor based on CO2 concentration.
- Parameters
events_file – file containing events
no_event_time_shift – number of seconds subtracted from start of event to define event when a window was closed
cls – list of clients
devs – list of devices
lat – latitude of a locality
lon – longitude of a locality
weather – object used to get information about weather
- Returns
dictionary of training data for predictor based on CO2 concentration
-
dm.models.predictor.generic_training_file.
training_file_t_h
(events_file: str, no_event_time_shift: int, cls, devs, lat, lon, weather)¶ It creates training data for predictor based on temperature and humidity.
- Parameters
events_file – file containing events
no_event_time_shift – number of seconds subtracted from start of event to define event when a window was closed
cls – list of clients
devs – list of devices
lat – latitude of a locality
lon – longitude of a locality
weather – object used to get information about weather
- Returns
dictionary of training data for predictor based on temperature and humidity
dm.models.predictor.generic_training_file_from_local_db module¶
Creates training datasets from local database.
-
dm.models.predictor.generic_training_file_from_local_db.
create_bins
(data, bins)¶ It creates required number of bins.
- Parameters
data – dictionary of data
bins – number of bins
- Returns
dictionary of data
-
dm.models.predictor.generic_training_file_from_local_db.
training_file_co2
(events_file: str, no_event_time_shift: int)¶ It creates training data for predictor based on CO2 concentration.
- Parameters
events_file – file containing events
no_event_time_shift – number of seconds subtracted from start of event to define event when a window was closed
- Returns
dictionary of training data for predictor based on CO2 concentration.
-
dm.models.predictor.generic_training_file_from_local_db.
training_file_t_h
(events_file: str, no_event_time_shift: int)¶ It creates training data for predictor based on temperature and humidity.
- Parameters
events_file – file containing events
no_event_time_shift – number of seconds subtracted from start of event to define event when a window was closed
- Returns
dictionary of training data for predictor based on temperature and humidity.