Getting started

Note

For a demonstration of usage for simple forecasting refer to Jupyter notebook simple_forecasting

Code snippets

Download a dataset

from cesnet_tszoo.datasets import CESNET_TimeSeries24
from cesnet_tszoo.utils.enums import SourceType, AgreggationType

dataset = CESNET_TimeSeries24.get_dataset("/some_directory/", source_type=SourceType.INSTITUTIONS, aggregation=AgreggationType.AGG_1_DAY, is_series_based=False)

Alternatively you can use load_benchmark.

from cesnet_tszoo.benchmarks import load_benchmark

benchmark = load_benchmark("SOME_BUILT_IN_IDENTIFIER", "/some_directory/")
dataset = benchmark.get_dataset()

This will create following directories:

• "/some_directory/tszoo"
  • "/some_directory/tszoo/annotations"
  • "/some_directory/tszoo/benchmarks"
  • "/some_directory/tszoo/configs"
  • "/some_directory/tszoo/databases"

Dataset will be downloaded to "/some_directory/tszoo/databases/CESNET_TimeSeries24/".

Enable logging

import logging

logging.basicConfig(
    level=logging.INFO,
    format="[%(asctime)s][%(name)s][%(levelname)s] - %(message)s")

Set up logging to get more information from the package.

Initialize dataset to create train, validation, and test sets

Using TimeBasedCesnetDataset dataset

from cesnet_tszoo.datasets import CESNET_TimeSeries24
from cesnet_tszoo.utils.enums import SourceType, AgreggationType
from cesnet_tszoo.configs import TimeBasedConfig

dataset = CESNET_TimeSeries24.get_dataset("/some_directory/", source_type=SourceType.INSTITUTIONS, aggregation=AgreggationType.AGG_1_DAY, is_series_based=False)
config = TimeBasedConfig(
    ts_ids=50, # number of randomly selected time series from dataset
    train_time_period=range(0, 100), 
    val_time_period=range(100, 150), 
    test_time_period=range(150, 250), 
    features_to_take=["n_flows", "n_packets"])
dataset.set_dataset_config_and_initialize(config)

Time-based datasets are configured with TimeBasedConfig. Can load data using:

• get_train_dataloader, get_val_dataloader, get_test_dataloader, get_test_other_dataloader, get_all_dataloader

• get_train_df, get_val_df, get_test_df, get_test_other_df, get_all_df

• get_train_numpy, get_val_numpy, get_test_numpy, get_test_other_numpy, get_all_numpy

Using SeriesBasedCesnetDataset dataset

from cesnet_tszoo.datasets import CESNET_TimeSeries24
from cesnet_tszoo.utils.enums import SourceType, AgreggationType
from cesnet_tszoo.configs import SeriesBasedConfig

dataset = CESNET_TimeSeries24.get_dataset("/some_directory/", source_type=SourceType.INSTITUTIONS, aggregation=AgreggationType.AGG_1_DAY, is_series_based=True)
config = SeriesBasedConfig(
    time_period=range(0, 250), 
    train_ts=100, # number of randomly selected time series from dataset
    val_ts=30, # number of randomly selected time series from dataset
    test_ts=20, # number of randomly selected time series from dataset
    features_to_take=["n_flows", "n_packets"])
dataset.set_dataset_config_and_initialize(config)

Series-based datasets are configured with SeriesBasedConfig. Can load data using:

• get_train_dataloader, get_val_dataloader, get_test_dataloader, get_all_dataloader

• get_train_df, get_val_df, get_test_df, get_all_df

• get_train_numpy, get_val_numpy, get_test_numpy, get_all_numpy

Using load_benchmark

from cesnet_tszoo.benchmarks import load_benchmark

benchmark = load_benchmark("SOME_BUILT_IN_IDENTIFIER", "/some_directory/")
dataset = benchmark.get_initialized_dataset()

Whether loaded dataset is series-based or time-based depends on the benchmark. What can be loaded corresponds to previous datasets.