import json
import logging
import math
import os
import pprint
import time
from pathlib import Path
from typing import Any, Dict, List, Literal, Optional, Tuple, Type, Union
import numpy as np
import pandas as pd
from autogluon.common.utils.deprecated_utils import Deprecated
from autogluon.common.utils.log_utils import add_log_to_file, set_logger_verbosity
from autogluon.common.utils.system_info import get_ag_system_info
from autogluon.common.utils.utils import check_saved_predictor_version, setup_outputdir
from autogluon.core.utils.decorators import apply_presets
from autogluon.core.utils.loaders import load_pkl, load_str
from autogluon.core.utils.savers import save_pkl, save_str
from autogluon.timeseries import __version__ as current_ag_version
from autogluon.timeseries.configs import TIMESERIES_PRESETS_CONFIGS
from autogluon.timeseries.dataset.ts_dataframe import ITEMID, TimeSeriesDataFrame
from autogluon.timeseries.learner import AbstractLearner, TimeSeriesLearner
from autogluon.timeseries.metrics import TimeSeriesScorer, check_get_evaluation_metric
from autogluon.timeseries.splitter import ExpandingWindowSplitter
from autogluon.timeseries.trainer import AbstractTimeSeriesTrainer
logger = logging.getLogger("autogluon.timeseries")
class TimeSeriesPredictorDeprecatedMixin:
"""Contains deprecated methods from TimeSeriesPredictor that shouldn't show up in API documentation."""
@Deprecated(min_version_to_warn="0.8.3", min_version_to_error="1.2", version_to_remove="1.2", new="evaluate")
def score(self, *args, **kwargs):
return self.evaluate(*args, **kwargs)
@Deprecated(min_version_to_warn="0.8.3", min_version_to_error="1.2", version_to_remove="1.2", new="model_best")
def get_model_best(self) -> str:
return self.model_best
@Deprecated(min_version_to_warn="0.8.3", min_version_to_error="1.2", version_to_remove="1.2", new="model_names")
def get_model_names(self) -> str:
return self.model_names()
[docs]
class TimeSeriesPredictor(TimeSeriesPredictorDeprecatedMixin):
"""AutoGluon ``TimeSeriesPredictor`` predicts future values of multiple related time series.
``TimeSeriesPredictor`` provides probabilistic (quantile) multi-step-ahead forecasts for univariate time series.
The forecast includes both the mean (i.e., conditional expectation of future values given the past), as well as the
quantiles of the forecast distribution, indicating the range of possible future outcomes.
``TimeSeriesPredictor`` fits both "global" deep learning models that are shared across all time series
(e.g., DeepAR, Transformer), as well as "local" statistical models that are fit to each individual time series
(e.g., ARIMA, ETS).
``TimeSeriesPredictor`` expects input data and makes predictions in the
:class:`~autogluon.timeseries.TimeSeriesDataFrame` format.
Parameters
----------
target : str, default = "target"
Name of column that contains the target values to forecast (i.e., numeric observations of the time series).
prediction_length : int, default = 1
The forecast horizon, i.e., How many time steps into the future the models should be trained to predict.
For example, if time series contain daily observations, setting ``prediction_length = 3`` will train
models that predict up to 3 days into the future from the most recent observation.
freq : str, optional
Frequency of the time series data (see `pandas documentation <https://pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases>`_
for available frequencies). For example, ``"D"`` for daily data or ``"h"`` for hourly data.
By default, the predictor will attempt to automatically infer the frequency from the data. This argument should
only be set in two cases:
1. The time series data has irregular timestamps, so frequency cannot be inferred automatically.
2. You would like to resample the original data at a different frequency (for example, convert hourly measurements into daily measurements).
If ``freq`` is provided when creating the predictor, all data passed to the predictor will be automatically
resampled at this frequency.
eval_metric : Union[str, TimeSeriesScorer], default = "WQL"
Metric by which predictions will be ultimately evaluated on future test data. AutoGluon tunes hyperparameters
in order to improve this metric on validation data, and ranks models (on validation data) according to this
metric.
Probabilistic forecast metrics (evaluated on quantile forecasts for the specified ``quantile_levels``):
- ``"SQL"``: scaled quantile loss
- ``"WQL"``: weighted quantile loss
Point forecast metrics (these are always evaluated on the ``"mean"`` column of the predictions):
- ``"MAE"``: mean absolute error
- ``"MAPE"``: mean absolute percentage error
- ``"MASE"``: mean absolute scaled error
- ``"MSE"``: mean squared error
- ``"RMSE"``: root mean squared error
- ``"RMSLE"``: root mean squared logarithmic error
- ``"RMSSE"``: root mean squared scaled error
- ``"SMAPE"``: "symmetric" mean absolute percentage error
- ``"WAPE"``: weighted absolute percentage error
For more information about these metrics, see :ref:`Forecasting Time Series - Evaluation Metrics <forecasting_metrics>`.
eval_metric_seasonal_period : int, optional
Seasonal period used to compute some evaluation metrics such as mean absolute scaled error (MASE). Defaults to
``None``, in which case the seasonal period is computed based on the data frequency.
known_covariates_names: List[str], optional
Names of the covariates that are known in advance for all time steps in the forecast horizon. These are also
known as dynamic features, exogenous variables, additional regressors or related time series. Examples of such
covariates include holidays, promotions or weather forecasts.
Currently, only numeric (float of integer dtype) are supported.
If ``known_covariates_names`` are provided, then:
- :meth:`~autogluon.timeseries.TimeSeriesPredictor.fit`, :meth:`~autogluon.timeseries.TimeSeriesPredictor.evaluate`, and :meth:`~autogluon.timeseries.TimeSeriesPredictor.leaderboard` will expect a data frame with columns listed in ``known_covariates_names`` (in addition to the ``target`` column).
- :meth:`~autogluon.timeseries.TimeSeriesPredictor.predict` will expect an additional keyword argument ``known_covariates`` containing the future values of the known covariates in ``TimeSeriesDataFrame`` format.
quantile_levels : List[float], optional
List of increasing decimals that specifies which quantiles should be estimated when making distributional
forecasts. Defaults to ``[0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]``.
path : str or pathlib.Path, optional
Path to the directory where models and intermediate outputs will be saved. Defaults to a timestamped folder
``AutogluonModels/ag-[TIMESTAMP]`` that will be created in the working directory.
verbosity : int, default = 2
Verbosity levels range from 0 to 4 and control how much information is printed to stdout. Higher levels
correspond to more detailed print statements, and ``verbosity=0`` suppresses output including warnings.
Verbosity 0 corresponds to Python's ERROR log level, where only error outputs will be logged. Verbosity 1 and 2
will additionally log warnings and info outputs, respectively. Verbosity 4 enables all logging output including
debug messages from AutoGluon and all logging in dependencies (GluonTS, PyTorch Lightning, AutoGluon-Tabular, etc.)
log_to_file: bool, default = True
Whether to save the logs into a file for later reference
log_file_path: Union[str, Path], default = "auto"
File path to save the logs.
If auto, logs will be saved under `predictor_path/logs/predictor_log.txt`.
Will be ignored if `log_to_file` is set to False
cache_predictions : bool, default = True
If True, the predictor will cache and reuse the predictions made by individual models whenever
:meth:`~autogluon.timeseries.TimeSeriesPredictor.predict`, :meth:`~autogluon.timeseries.TimeSeriesPredictor.leaderboard`,
or :meth:`~autogluon.timeseries.TimeSeriesPredictor.evaluate` methods are called. This allows to significantly
speed up these methods. If False, caching will be disabled. You can set this argument to False to reduce disk
usage at the cost of longer prediction times.
label : str, optional
Alias for :attr:`target`.
"""
predictor_file_name = "predictor.pkl"
_predictor_version_file_name = "version.txt"
_predictor_log_file_name = "predictor_log.txt"
[docs]
def __init__(
self,
target: Optional[str] = None,
known_covariates_names: Optional[List[str]] = None,
prediction_length: int = 1,
freq: str = None,
eval_metric: Union[str, TimeSeriesScorer, None] = None,
eval_metric_seasonal_period: Optional[int] = None,
path: Optional[Union[str, Path]] = None,
verbosity: int = 2,
log_to_file: bool = True,
log_file_path: Union[str, Path] = "auto",
quantile_levels: Optional[List[float]] = None,
cache_predictions: bool = True,
learner_type: Optional[Type[AbstractLearner]] = None,
learner_kwargs: Optional[dict] = None,
label: Optional[str] = None,
**kwargs,
):
self.verbosity = verbosity
set_logger_verbosity(self.verbosity, logger=logger)
self.path = setup_outputdir(path)
self._setup_log_to_file(log_to_file=log_to_file, log_file_path=log_file_path)
self.cache_predictions = cache_predictions
if target is not None and label is not None:
raise ValueError("Both `label` and `target` are specified. Please specify at most one of these arguments.")
self.target = target or label or "target"
if known_covariates_names is None:
known_covariates_names = []
if isinstance(known_covariates_names, str):
known_covariates_names = [known_covariates_names]
if not all(isinstance(name, str) for name in known_covariates_names):
raise ValueError(
"known_covariates_names must be a list of strings (names of columns that are known at prediction time)."
)
if self.target in known_covariates_names:
raise ValueError(f"Target column {self.target} cannot be one of the known covariates.")
self.known_covariates_names = list(known_covariates_names)
self.prediction_length = int(prediction_length)
# For each validation fold, all time series in training set must have length >= _min_train_length
self._min_train_length = max(self.prediction_length + 1, 5)
self.freq = freq
if self.freq is not None:
# Standardize frequency string (e.g., "T" -> "min", "Y" -> "YE")
std_freq = pd.tseries.frequencies.to_offset(self.freq).freqstr
if std_freq != str(self.freq):
logger.info(f"Frequency '{self.freq}' stored as '{std_freq}'")
self.freq = std_freq
self.eval_metric = check_get_evaluation_metric(eval_metric)
self.eval_metric_seasonal_period = eval_metric_seasonal_period
if quantile_levels is None:
quantile_levels = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]
self.quantile_levels = sorted(quantile_levels)
if learner_kwargs is None:
learner_kwargs = {}
learner_kwargs = learner_kwargs.copy()
learner_kwargs.update(
dict(
path_context=self.path,
eval_metric=eval_metric,
eval_metric_seasonal_period=eval_metric_seasonal_period,
target=self.target,
known_covariates_names=self.known_covariates_names,
prediction_length=self.prediction_length,
quantile_levels=self.quantile_levels,
cache_predictions=self.cache_predictions,
)
)
# Using `TimeSeriesLearner` as default argument breaks doc generation with Sphnix
if learner_type is None:
learner_type = TimeSeriesLearner
self._learner: AbstractLearner = learner_type(**learner_kwargs)
self._learner_type = type(self._learner)
if "ignore_time_index" in kwargs:
raise TypeError(
"`ignore_time_index` argument to TimeSeriesPredictor.__init__() has been deprecated.\n"
"If your data has irregular timestamps, please either 1) specify the desired regular frequency when "
"creating the predictor as `TimeSeriesPredictor(freq=...)` or 2) manually convert timestamps to "
"regular frequency with `data.convert_frequency(freq=...)`."
)
if len(kwargs) > 0:
for key in kwargs:
raise TypeError(f"TimeSeriesPredictor.__init__() got an unexpected keyword argument '{key}'")
@property
def _trainer(self) -> AbstractTimeSeriesTrainer:
return self._learner.load_trainer() # noqa
def _setup_log_to_file(self, log_to_file: bool, log_file_path: Union[str, Path]) -> None:
if log_to_file:
if log_file_path == "auto":
log_file_path = os.path.join(self.path, "logs", self._predictor_log_file_name)
log_file_path = os.path.abspath(os.path.normpath(log_file_path))
os.makedirs(os.path.dirname(log_file_path), exist_ok=True)
add_log_to_file(log_file_path)
def _to_data_frame(
self,
data: Union[TimeSeriesDataFrame, pd.DataFrame, Path, str],
name: str = "data",
) -> "TimeSeriesDataFrame":
if isinstance(data, TimeSeriesDataFrame):
return data
elif isinstance(data, (pd.DataFrame, Path, str)):
try:
data = TimeSeriesDataFrame(data)
except:
raise ValueError(
f"Provided {name} of type {type(data)} cannot be automatically converted to a TimeSeriesDataFrame."
)
return data
else:
raise TypeError(
f"{name} must be a TimeSeriesDataFrame, pandas.DataFrame, pathlib.Path or string (path to data) "
f"but received an object of type {type(data)}."
)
def _check_and_prepare_data_frame(
self,
data: Union[TimeSeriesDataFrame, pd.DataFrame, Path, str],
name: str = "data",
) -> TimeSeriesDataFrame:
"""Ensure that TimeSeriesDataFrame has a sorted index and a valid frequency.
If self.freq is None, then self.freq of the predictor will be set to the frequency of the data.
Parameters
----------
data : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]
Data as a data frame or path to file storing the data.
name : str
Name of the data that will be used in log messages (e.g., 'train_data', 'tuning_data', or 'data').
Returns
-------
df : TimeSeriesDataFrame
Preprocessed data in TimeSeriesDataFrame format.
"""
df = self._to_data_frame(data, name=name)
df = df.astype({self.target: "float64"})
# MultiIndex.is_monotonic_increasing checks if index is sorted by ["item_id", "timestamp"]
if not df.index.is_monotonic_increasing:
df = df.sort_index()
df._cached_freq = None # in case frequency was incorrectly cached as IRREGULAR_TIME_INDEX_FREQSTR
# Ensure that data has a regular frequency that matches the predictor frequency
if self.freq is None:
if df.freq is None:
raise ValueError(
f"Frequency of {name} is not provided and cannot be inferred. Please set the expected data "
f"frequency when creating the predictor with `TimeSeriesPredictor(freq=...)` or ensure that "
f"the data has a regular time index with `{name}.convert_frequency(freq=...)`"
)
else:
self.freq = df.freq
logger.info(f"Inferred time series frequency: '{df.freq}'")
else:
if df.freq != self.freq:
logger.warning(f"{name} with frequency '{df.freq}' has been resampled to frequency '{self.freq}'.")
df = df.convert_frequency(freq=self.freq)
return df
def _check_data_for_evaluation(self, data: TimeSeriesDataFrame, name: str = "data"):
"""Make sure that provided evaluation data includes both historic and future time series values."""
if data.num_timesteps_per_item().min() <= self.prediction_length:
raise ValueError(
f"Cannot reserve last prediction_length={self.prediction_length} time steps for evaluation in some "
f"time series in {name}. Please make sure that {name} includes both historic and future data, and that"
f"all time series have length > prediction_length (at least {self.prediction_length + 1})"
)
def _get_dataset_stats(self, data: TimeSeriesDataFrame) -> str:
ts_lengths = data.num_timesteps_per_item()
median_length = ts_lengths.median()
min_length = ts_lengths.min()
max_length = ts_lengths.max()
missing_value_fraction = data[self.target].isna().mean()
if missing_value_fraction > 0:
missing_value_fraction_str = f" (NaN fraction={missing_value_fraction:.1%})"
else:
missing_value_fraction_str = ""
return (
f"{len(data)} rows{missing_value_fraction_str}, {data.num_items} time series. "
f"Median time series length is {median_length:.0f} (min={min_length}, max={max_length}). "
)
def _reduce_num_val_windows_if_necessary(
self,
train_data: TimeSeriesDataFrame,
original_num_val_windows: int,
val_step_size: int,
) -> int:
"""Adjust num_val_windows based on the length of time series in train_data.
Chooses num_val_windows such that TS with median length is long enough to perform num_val_windows validations
(at least 1, at most `original_num_val_windows`).
In other words, find largest `num_val_windows` that satisfies
median_length >= min_train_length + prediction_length + (num_val_windows - 1) * val_step_size
"""
median_length = train_data.num_timesteps_per_item().median()
num_val_windows_for_median_ts = int(
(median_length - self._min_train_length - self.prediction_length) // val_step_size + 1
)
new_num_val_windows = min(original_num_val_windows, max(1, num_val_windows_for_median_ts))
if new_num_val_windows < original_num_val_windows:
logger.warning(
f"Time series in train_data are too short for chosen num_val_windows={original_num_val_windows}. "
f"Reducing num_val_windows to {new_num_val_windows}."
)
return new_num_val_windows
def _filter_useless_train_data(
self,
train_data: TimeSeriesDataFrame,
num_val_windows: int,
val_step_size: int,
) -> Tuple[TimeSeriesDataFrame, Optional[TimeSeriesDataFrame]]:
"""Remove time series from train_data that either contain all NaNs or are too short for chosen settings.
This method ensures that 1) no time series consist of all NaN values and 2) for each validation fold, all train
series have length >= max(prediction_length + 1, 5).
In other words, this method removes from train_data all time series with only NaN values or length less than
min_train_length + prediction_length + (num_val_windows - 1) * val_step_size
"""
min_length = self._min_train_length + self.prediction_length + (num_val_windows - 1) * val_step_size
train_lengths = train_data.num_timesteps_per_item()
too_short_items = train_lengths.index[train_lengths < min_length]
if len(too_short_items) > 0:
logger.info(
f"\tRemoving {len(too_short_items)} short time series from train_data. Only series with length "
f">= {min_length} will be used for training."
)
train_data = train_data.query("item_id not in @too_short_items")
all_nan_items = train_data.item_ids[train_data[self.target].isna().groupby(ITEMID, sort=False).all()]
if len(all_nan_items) > 0:
logger.info(f"\tRemoving {len(all_nan_items)} time series consisting of only NaN values from train_data.")
train_data = train_data.query("item_id not in @all_nan_items")
if len(too_short_items) or len(all_nan_items):
logger.info(f"\tAfter filtering, train_data has {self._get_dataset_stats(train_data)}")
if len(train_data) == 0:
raise ValueError(
f"At least some time series in train_data must have >= {min_length} observations. Please provide "
f"longer time series as train_data or reduce prediction_length, num_val_windows, or val_step_size."
)
return train_data
[docs]
@apply_presets(TIMESERIES_PRESETS_CONFIGS)
def fit(
self,
train_data: Union[TimeSeriesDataFrame, pd.DataFrame, Path, str],
tuning_data: Optional[Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]] = None,
time_limit: Optional[int] = None,
presets: Optional[str] = None,
hyperparameters: Dict[Union[str, Type], Any] = None,
hyperparameter_tune_kwargs: Optional[Union[str, Dict]] = None,
excluded_model_types: Optional[List[str]] = None,
num_val_windows: int = 1,
val_step_size: Optional[int] = None,
refit_every_n_windows: int = 1,
refit_full: bool = False,
enable_ensemble: bool = True,
skip_model_selection: bool = False,
random_seed: Optional[int] = 123,
verbosity: Optional[int] = None,
) -> "TimeSeriesPredictor":
"""Fit probabilistic forecasting models to the given time series dataset.
Parameters
----------
train_data : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]
Training data in the :class:`~autogluon.timeseries.TimeSeriesDataFrame` format.
Time series with length ``<= (num_val_windows + 1) * prediction_length`` will be ignored during training.
See :attr:`num_val_windows` for details.
If ``known_covariates_names`` were specified when creating the predictor, ``train_data`` must include the
columns listed in ``known_covariates_names`` with the covariates values aligned with the target time series.
The known covariates must have a numeric (float or integer) dtype.
Columns of ``train_data`` except ``target`` and those listed in ``known_covariates_names`` will be
interpreted as ``past_covariates`` - covariates that are known only in the past.
If ``train_data`` has static features (i.e., ``train_data.static_features`` is a pandas DataFrame), the
predictor will interpret columns with ``int`` and ``float`` dtypes as continuous (real-valued) features,
columns with ``object`` and ``str`` dtypes as categorical features, and will ignore the rest of columns.
For example, to ensure that column "store_id" with dtype ``int`` is interpreted as a category,
we need to change its type to ``category``::
data.static_features["store_id"] = data.static_features["store_id"].astype("category")
If provided data is a path or a pandas.DataFrame, AutoGluon will attempt to automatically convert it to a
``TimeSeriesDataFrame``.
tuning_data : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str], optional
Data reserved for model selection and hyperparameter tuning, rather than training individual models. Also
used to compute the validation scores. Note that only the last ``prediction_length`` time steps of each
time series are used for computing the validation score.
If ``tuning_data`` is provided, multi-window backtesting on training data will be disabled, the
:attr:`num_val_windows` will be set to ``0``, and :attr:`refit_full` will be set to ``False``.
Leaving this argument empty and letting AutoGluon automatically generate the validation set from
``train_data`` is a good default.
If ``known_covariates_names`` were specified when creating the predictor, ``tuning_data`` must also include
the columns listed in ``known_covariates_names`` with the covariates values aligned with the target time
series.
If ``train_data`` has past covariates or static features, ``tuning_data`` must have also include them (with
same columns names and dtypes).
If provided data is a path or a pandas.DataFrame, AutoGluon will attempt to automatically convert it to a
``TimeSeriesDataFrame``.
time_limit : int, optional
Approximately how long :meth:`~autogluon.timeseries.TimeSeriesPredictor.fit` will run (wall-clock time in
seconds). If not specified, :meth:`~autogluon.timeseries.TimeSeriesPredictor.fit` will run until all models
have completed training.
presets : str, optional
Optional preset configurations for various arguments in
:meth:`~autogluon.timeseries.TimeSeriesPredictor.fit`.
Can significantly impact predictive accuracy, memory footprint, inference latency of trained models,
and various other properties of the returned predictor. It is recommended to specify presets and avoid
specifying most other :meth:`~autogluon.timeseries.TimeSeriesPredictor.fit` arguments or model
hyperparameters prior to becoming familiar with AutoGluon. For example, set ``presets="high_quality"``
to get a high-accuracy predictor, or set ``presets="fast_training"`` to quickly get the results.
Any user-specified arguments in :meth:`~autogluon.timeseries.TimeSeriesPredictor.fit` will
override the values used by presets.
Available presets:
- ``"fast_training"``: fit simple statistical models (``ETS``, ``Theta``, ``Naive``, ``SeasonalNaive``) + fast tree-based models ``RecursiveTabular``
and ``DirectTabular``. These models are fast to train but may not be very accurate.
- ``"medium_quality"``: all models mentioned above + deep learning model ``TemporalFusionTransformer``. Default setting that produces good forecasts
with reasonable training time.
- ``"high_quality"``: All ML models available in AutoGluon + additional statistical models (``NPTS``, ``AutoETS``, ``AutoARIMA``, ``CrostonSBA``,
``DynamicOptimizedTheta``). Much more accurate than ``medium_quality``, but takes longer to train.
- ``"best_quality"``: Same models as in ``"high_quality"``, but performs validation with multiple backtests. Usually better than ``high_quality``, but takes even longer to train.
Available presets with the `Chronos <https://github.com/amazon-science/chronos-forecasting>`_ model:
- ``"chronos_{model_size}"``: where model size is one of ``tiny,mini,small,base,large``. Uses the Chronos pretrained model for zero-shot forecasting.
See the documentation for ``ChronosModel`` or see `Hugging Face <https://huggingface.co/collections/amazon/chronos-models-65f1791d630a8d57cb718444>`_ for more information.
Note that a GPU is required for model sizes ``small``, ``base`` and ``large``.
- ``"chronos"``: alias for ``"chronos_small"``.
- ``"chronos_ensemble"``: builds an ensemble of seasonal naive, tree-based and deep learning models with fast inference
and ``"chronos_small"``.
- ``"chronos_large_ensemble"``: builds an ensemble of seasonal naive, tree-based and deep learning models
with fast inference and ``"chronos_large"``.
Details for these presets can be found in ``autogluon/timeseries/configs/presets_configs.py``. If not
provided, user-provided values for ``hyperparameters`` and ``hyperparameter_tune_kwargs`` will be used
(defaulting to their default values specified below).
hyperparameters : str or dict, optional
Determines what models are trained and what hyperparameters are used by each model.
If str is passed, will use a preset hyperparameter configuration defined in
``autogluon/timeseries/trainer/models/presets.py``. Supported values are ``"default"``, ``"light"`` and
``"very_light"``.
If dict is provided, the keys are strings or types that indicate which models to train. Each value is
itself a dict containing hyperparameters for each of the trained models, or a list of such dicts. Any
omitted hyperparameters not specified here will be set to default. For example::
predictor.fit(
...
hyperparameters={
"DeepAR": {},
"Theta": [
{"decomposition_type": "additive"},
{"seasonal_period": 1},
],
}
)
The above example will train three models:
* ``DeepAR`` with default hyperparameters
* ``Theta`` with additive seasonal decomposition (all other parameters set to their defaults)
* ``Theta`` with seasonality disabled (all other parameters set to their defaults)
Full list of available models and their hyperparameters is provided in :ref:`Forecasting Time Series - Model Zoo <forecasting_model_zoo>`.
The hyperparameters for each model can be fixed values (as shown above), or search spaces over which
hyperparameter optimization is performed. A search space should only be provided when
``hyperparameter_tune_kwargs`` is given (i.e., hyperparameter-tuning is utilized). For example::
from autogluon.common import space
predictor.fit(
...
hyperparameters={
"DeepAR": {
"hidden_size": space.Int(20, 100),
"dropout_rate": space.Categorical(0.1, 0.3),
},
},
hyperparameter_tune_kwargs="auto",
)
In the above example, multiple versions of the DeepAR model with different values of the parameters
"hidden_size" and "dropout_rate" will be trained.
hyperparameter_tune_kwargs : str or dict, optional
Hyperparameter tuning strategy and kwargs (for example, how many HPO trials to run).
If None, then hyperparameter tuning will not be performed.
If type is ``str``, then this argument specifies a preset.
Valid preset values:
* "auto": Performs HPO via bayesian optimization search on GluonTS-backed neural forecasting models and
random search on other models using local scheduler.
* "random": Performs HPO via random search.
You can also provide a dict to specify searchers and schedulers
Valid keys:
* "num_trials": How many HPO trials to run
* "scheduler": Which scheduler to use. Valid values:
* "local": Local scheduler that schedules trials FIFO
* "searcher": Which searching algorithm to use. Valid values:
* "local_random": Uses the "random" searcher
* "random": Perform random search
* "bayes": Perform HPO with HyperOpt on GluonTS-backed models via Ray tune. Perform random search on other models.
* "auto": alias for "bayes"
The "scheduler" and "searcher" key are required when providing a dict.
Example::
predictor.fit(
...
hyperparameter_tune_kwargs={
"num_trials": 5,
"searcher": "auto",
"scheduler": "local",
},
)
excluded_model_types: List[str], optional
Banned subset of model types to avoid training during ``fit()``, even if present in ``hyperparameters``.
For example, the following code will train all models included in the ``high_quality`` presets except ``DeepAR``::
predictor.fit(
...,
presets="high_quality",
excluded_model_types=["DeepAR"],
)
num_val_windows : int, default = 1
Number of backtests done on ``train_data`` for each trained model to estimate the validation performance.
If ``num_val_windows > 1`` is provided, this value may be automatically reduced to ensure that the majority
of time series in ``train_data`` are long enough for the chosen number of backtests.
Increasing this parameter increases the training time roughly by a factor of ``num_val_windows // refit_every_n_windows``.
See :attr:`refit_every_n_windows` and :attr:`val_step_size`: for details.
For example, for ``prediction_length=2``, ``num_val_windows=3`` and ``val_step_size=1`` the folds are::
|-------------------|
| x x x x x y y - - |
| x x x x x x y y - |
| x x x x x x x y y |
where ``x`` are the train time steps and ``y`` are the validation time steps.
This argument has no effect if ``tuning_data`` is provided.
val_step_size : int or None, default = None
Step size between consecutive validation windows. If set to ``None``, defaults to ``prediction_length``
provided when creating the predictor.
This argument has no effect if ``tuning_data`` is provided.
refit_every_n_windows: int or None, default = 1
When performing cross validation, each model will be retrained every ``refit_every_n_windows`` validation
windows, where the number of validation windows is specified by `num_val_windows`. Note that in the
default setting where `num_val_windows=1`, this argument has no effect.
If set to ``None``, models will only be fit once for the first (oldest) validation window. By default,
`refit_every_n_windows=1`, i.e., all models will be refit for each validation window.
refit_full : bool, default = False
If True, after training is complete, AutoGluon will attempt to re-train all models using all of training
data (including the data initially reserved for validation). This argument has no effect if ``tuning_data``
is provided.
enable_ensemble : bool, default = True
If True, the ``TimeSeriesPredictor`` will fit a simple weighted ensemble on top of the models specified via
``hyperparameters``.
skip_model_selection : bool, default = False
If True, predictor will not compute the validation score. For example, this argument is useful if we want
to use the predictor as a wrapper for a single pre-trained model. If set to True, then the ``hyperparameters``
dict must contain exactly one model without hyperparameter search spaces or an exception will be raised.
random_seed : int or None, default = 123
If provided, fixes the seed of the random number generator for all models. This guarantees reproducible
results for most models (except those trained on GPU because of the non-determinism of GPU operations).
verbosity : int, optional
If provided, overrides the ``verbosity`` value used when creating the ``TimeSeriesPredictor``. See
documentation for :class:`~autogluon.timeseries.TimeSeriesPredictor` for more details.
"""
time_start = time.time()
if self._learner.is_fit:
raise AssertionError("Predictor is already fit! To fit additional models create a new `Predictor`.")
if verbosity is None:
verbosity = self.verbosity
set_logger_verbosity(verbosity, logger=logger)
logger.info("Beginning AutoGluon training..." + (f" Time limit = {time_limit}s" if time_limit else ""))
logger.info(f"AutoGluon will save models to '{self.path}'")
logger.info(get_ag_system_info(path=self.path, include_gpu_count=True))
if hyperparameters is None:
hyperparameters = "default"
fit_args = dict(
prediction_length=self.prediction_length,
target=self.target,
known_covariates_names=self.known_covariates_names,
eval_metric=self.eval_metric,
eval_metric_seasonal_period=self.eval_metric_seasonal_period,
quantile_levels=self.quantile_levels,
freq=self.freq,
time_limit=time_limit,
hyperparameters=hyperparameters,
hyperparameter_tune_kwargs=hyperparameter_tune_kwargs,
excluded_model_types=excluded_model_types,
num_val_windows=num_val_windows,
val_step_size=val_step_size,
refit_every_n_windows=refit_every_n_windows,
refit_full=refit_full,
skip_model_selection=skip_model_selection,
enable_ensemble=enable_ensemble,
random_seed=random_seed,
verbosity=verbosity,
)
if presets is not None:
logger.info(f"Setting presets to: {presets}")
logger.info("\nFitting with arguments:")
logger.info(f"{pprint.pformat({k: v for k, v in fit_args.items() if v is not None})}\n")
train_data = self._check_and_prepare_data_frame(train_data, name="train_data")
logger.info(f"Provided train_data has {self._get_dataset_stats(train_data)}")
if val_step_size is None:
val_step_size = self.prediction_length
if num_val_windows > 0:
num_val_windows = self._reduce_num_val_windows_if_necessary(
train_data, original_num_val_windows=num_val_windows, val_step_size=val_step_size
)
if tuning_data is not None:
tuning_data = self._check_and_prepare_data_frame(tuning_data, name="tuning_data")
self._check_data_for_evaluation(tuning_data, name="tuning_data")
logger.info(f"Provided tuning_data has {self._get_dataset_stats(tuning_data)}")
# TODO: Use num_val_windows to perform multi-window backtests on tuning_data
if num_val_windows > 0:
logger.warning(
"\tSetting num_val_windows = 0 (disabling backtesting on train_data) because tuning_data is provided."
)
num_val_windows = 0
if num_val_windows == 0 and tuning_data is None:
raise ValueError("Please set num_val_windows >= 1 or provide custom tuning_data")
if num_val_windows <= 1 and refit_every_n_windows > 1:
logger.warning(
f"\trefit_every_n_windows provided as {refit_every_n_windows} but num_val_windows is set to {num_val_windows}."
" Refit_every_n_windows will have no effect."
)
if not skip_model_selection:
train_data = self._filter_useless_train_data(
train_data, num_val_windows=num_val_windows, val_step_size=val_step_size
)
val_splitter = ExpandingWindowSplitter(
prediction_length=self.prediction_length, num_val_windows=num_val_windows, val_step_size=val_step_size
)
time_left = None if time_limit is None else time_limit - (time.time() - time_start)
self._learner.fit(
train_data=train_data,
val_data=tuning_data,
hyperparameters=hyperparameters,
hyperparameter_tune_kwargs=hyperparameter_tune_kwargs,
excluded_model_types=excluded_model_types,
time_limit=time_left,
verbosity=verbosity,
val_splitter=val_splitter,
refit_every_n_windows=refit_every_n_windows,
skip_model_selection=skip_model_selection,
enable_ensemble=enable_ensemble,
random_seed=random_seed,
)
if refit_full:
if tuning_data is None:
self.refit_full()
else:
logger.warning("Skipping `refit_full` because custom `tuning_data` was provided during `fit`.")
self.save()
return self
[docs]
def model_names(self) -> List[str]:
"""Returns the list of model names trained by this predictor object."""
return self._trainer.get_model_names()
[docs]
def predict(
self,
data: Union[TimeSeriesDataFrame, pd.DataFrame, Path, str],
known_covariates: Optional[Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]] = None,
model: Optional[str] = None,
use_cache: bool = True,
random_seed: Optional[int] = 123,
) -> TimeSeriesDataFrame:
"""Return quantile and mean forecasts for the given dataset, starting from the end of each time series.
Parameters
----------
data : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]
Time series data to forecast with.
If ``known_covariates_names`` were specified when creating the predictor, ``data`` must include the columns
listed in ``known_covariates_names`` with the covariates values aligned with the target time series.
If ``train_data`` used to train the predictor contained past covariates or static features, then ``data``
must also include them (with same column names and dtypes).
If provided data is an instance of pandas DataFrame, AutoGluon will attempt to automatically convert it
to a ``TimeSeriesDataFrame``.
known_covariates : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str], optional
If ``known_covariates_names`` were specified when creating the predictor, it is necessary to provide the
values of the known covariates for each time series during the forecast horizon. That is:
- The columns must include all columns listed in ``known_covariates_names``
- The ``item_id`` index must include all item ids present in ``data``
- The ``timestamp`` index must include the values for ``prediction_length`` many time steps into the future from the end of each time series in ``data``
See example below.
model : str, optional
Name of the model that you would like to use for prediction. By default, the best model during training
(with highest validation score) will be used.
random_seed : int or None, default = 123
If provided, fixes the seed of the random number generator for all models. This guarantees reproducible
results for most models (except those trained on GPU because of the non-determinism of GPU operations).
use_cache : bool, default = True
If True, will attempt to use the cached predictions. If False, cached predictions will be ignored.
This argument is ignored if ``cache_predictions`` was set to False when creating the ``TimeSeriesPredictor``.
Examples
--------
>>> print(data)
target promotion price
item_id timestamp
A 2020-01-05 20 0 19.9
2020-01-06 40 1 9.9
2020-01-07 32 0 15.0
B 2020-03-01 13 0 5.0
2020-03-02 44 1 2.9
2020-03-03 72 1 2.9
>>> predictor = TimeSeriesPredictor(prediction_length=2, known_covariates_names=["promotion", "price"]).fit(data)
>>> print(future_known_covariates)
promotion price
item_id timestamp
A 2020-01-08 1 12.9
2020-01-09 1 12.9
B 2020-03-04 0 5.0
2020-03-05 0 7.0
>>> predictor.predict(data, known_covariates=future_known_covariates)
mean
item_id timestamp
A 2020-01-08 30.2
2020-01-09 27.0
B 2020-03-04 17.1
2020-03-05 8.3
"""
# Save original item_id order to return predictions in the same order as input data
data = self._to_data_frame(data)
original_item_id_order = data.item_ids
data = self._check_and_prepare_data_frame(data)
if known_covariates is not None:
known_covariates = self._to_data_frame(known_covariates)
predictions = self._learner.predict(
data,
known_covariates=known_covariates,
model=model,
use_cache=use_cache,
random_seed=random_seed,
)
return predictions.reindex(original_item_id_order, level=ITEMID)
[docs]
def evaluate(
self,
data: Union[TimeSeriesDataFrame, pd.DataFrame, Path, str],
model: Optional[str] = None,
metrics: Optional[Union[str, TimeSeriesScorer, List[Union[str, TimeSeriesScorer]]]] = None,
display: bool = False,
use_cache: bool = True,
) -> Dict[str, float]:
"""Evaluate the forecast accuracy for given dataset.
This method measures the forecast accuracy using the last ``self.prediction_length`` time steps of each time
series in ``data`` as a hold-out set.
Parameters
----------
data : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]
The data to evaluate the best model on. The last ``prediction_length`` time steps of each time series in
``data`` will be held out for prediction and forecast accuracy will be calculated on these time steps.
Must include both historic and future data (i.e., length of all time series in ``data`` must be at least
``prediction_length + 1``).
If ``known_covariates_names`` were specified when creating the predictor, ``data`` must include the columns
listed in ``known_covariates_names`` with the covariates values aligned with the target time series.
If ``train_data`` used to train the predictor contained past covariates or static features, then ``data``
must also include them (with same column names and dtypes).
If provided data is an instance of pandas DataFrame, AutoGluon will attempt to automatically convert it
to a ``TimeSeriesDataFrame``.
model : str, optional
Name of the model that you would like to evaluate. By default, the best model during training
(with highest validation score) will be used.
metrics : str, TimeSeriesScorer or List[Union[str, TimeSeriesScorer]], optional
Metric or a list of metrics to compute scores with. Defaults to ``self.eval_metric``. Supports both
metric names as strings and custom metrics based on TimeSeriesScorer.
display : bool, default = False
If True, the scores will be printed.
use_cache : bool, default = True
If True, will attempt to use the cached predictions. If False, cached predictions will be ignored.
This argument is ignored if ``cache_predictions`` was set to False when creating the ``TimeSeriesPredictor``.
Returns
-------
scores_dict : Dict[str, float]
Dictionary where keys = metrics, values = performance along each metric. For consistency, error metrics
will have their signs flipped to obey this convention. For example, negative MAPE values will be reported.
To get the ``eval_metric`` score, do ``output[predictor.eval_metric.name]``.
"""
data = self._check_and_prepare_data_frame(data)
self._check_data_for_evaluation(data)
scores_dict = self._learner.evaluate(data, model=model, metrics=metrics, use_cache=use_cache)
if display:
logger.info("Evaluations on test data:")
logger.info(json.dumps(scores_dict, indent=4))
return scores_dict
[docs]
def feature_importance(
self,
data: Optional[Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]] = None,
model: Optional[str] = None,
metric: Optional[Union[str, TimeSeriesScorer]] = None,
features: Optional[List[str]] = None,
time_limit: Optional[float] = None,
method: Literal["naive", "permutation"] = "permutation",
subsample_size: int = 50,
num_iterations: Optional[int] = None,
random_seed: Optional[int] = 123,
relative_scores: bool = False,
include_confidence_band: bool = True,
confidence_level: float = 0.99,
):
"""
Calculates feature importance scores for the given model via replacing each feature by a shuffled version of the same feature
(also known as permutation feature importance) or by assigning a constant value representing the median or mode of the feature,
and computing the relative decrease in the model's predictive performance.
A feature's importance score represents the performance drop that results when the model makes predictions on a perturbed copy
of the data where this feature's values have been randomly shuffled across rows. A feature score of 0.01 would indicate that the
predictive performance dropped by 0.01 when the feature was randomly shuffled or replaced. The higher the score a feature has,
the more important it is to the model's performance.
If a feature has a negative score, this means that the feature is likely harmful to the final model, and a model trained with
the feature removed would be expected to achieve a better predictive performance. Note that calculating feature importance can
be a computationally expensive process, particularly if the model uses many features. In many cases, this can take longer than
the original model training. Roughly, this will equal to the number of features in the data multiplied by ``num_iterations``
(or, 1 when ``method="naive"``) and time taken when ``evaluate()`` is called on a dataset with ``subsample_size``.
Parameters
----------
data : TimeSeriesDataFrame, pd.DataFrame, Path or str, optional
The data to evaluate feature importances on. The last ``prediction_length`` time steps of the data set, for each
item, will be held out for prediction and forecast accuracy will be calculated on these time steps.
More accurate feature importances will be obtained from new data that was held-out during ``fit()``.
If ``known_covariates_names`` were specified when creating the predictor, ``data`` must include the columns
listed in ``known_covariates_names`` with the covariates values aligned with the target time series.
This data must contain the label column with the same column name as specified during ``fit()``.
If ``train_data`` used to train the predictor contained past covariates or static features, then ``data``
must also include them (with same column names and dtypes).
If provided data is an instance of pandas DataFrame, AutoGluon will attempt to automatically convert it
to a ``TimeSeriesDataFrame``. If str or Path is passed, ``data`` will be loaded using the str value as the file path.
If ``data`` is not provided, then validation (tuning) data provided during training (or the held out data used for
validation if ``tuning_data`` was not explicitly provided ``fit()``) will be used.
model : str, optional
Name of the model that you would like to evaluate. By default, the best model during training
(with highest validation score) will be used.
metric : str or TimeSeriesScorer, optional
Metric to be used for computing feature importance. If None, the ``eval_metric`` specified during initialization of
the ``TimeSeriesPredictor`` will be used.
features : List[str], optional
List of feature names that feature importances are calculated for and returned. By default, all feature importances
will be returned.
method : {"permutation", "naive"}, default = "permutation"
Method to be used for computing feature importance.
* ``naive``: computes feature importance by replacing the values of each feature by a constant value and computing
feature importances as the relative improvement in the evaluation metric. The constant value is the median for
real-valued features and the mode for categorical features, for both covariates and static features, obtained from the
feature values in ``data`` provided.
* ``permutation``: computes feature importance by naively shuffling the values of the feature across different items
and time steps. Each feature is shuffled for ``num_iterations`` times and feature importances are computed as the
relative improvement in the evaluation metric. Refer to https://explained.ai/rf-importance/ for an explanation of
permutation importance.
subsample_size : int, default = 50
The number of items to sample from `data` when computing feature importance. Larger values increase the accuracy of
the feature importance scores. Runtime linearly scales with `subsample_size`.
time_limit : float, optional
Time in seconds to limit the calculation of feature importance. If None, feature importance will calculate without early stopping.
If ``method="permutation"``, a minimum of 1 full shuffle set will always be evaluated. If a shuffle set evaluation takes longer than
``time_limit``, the method will take the length of a shuffle set evaluation to return regardless of the `time_limit`.
num_iterations : int, optional
The number of different iterations of the data that are evaluated. If ``method="permutation"``, this will be interpreted
as the number of shuffle sets (equivalent to ``num_shuffle_sets`` in :meth:`TabularPredictor.feature_importance`). If ``method="naive"``, the
constant replacement approach is repeated for ``num_iterations`` times, and a different subsample of data (of size ``subsample_size``) will
be taken in each iteration.
Default is 1 for ``method="naive"`` and 5 for ``method="permutation"``. The value will be ignored if ``method="naive"`` and the subsample
size is greater than the number of items in ``data`` as additional iterations will be redundant.
Larger values will increase the quality of the importance evaluation.
It is generally recommended to increase ``subsample_size`` before increasing ``num_iterations``.
Runtime scales linearly with ``num_iterations``.
random_seed : int or None, default = 123
If provided, fixes the seed of the random number generator for all models. This guarantees reproducible
results for feature importance.
relative_scores : bool, default = False
By default, this method will return expected average *absolute* improvement in the eval metric due to the feature. If True, then
the statistics will be computed over the *relative* (percentage) improvements.
include_confidence_band: bool, default = True
If True, returned DataFrame will include two additional columns specifying confidence interval for the true underlying importance value of
each feature. Increasing ``subsample_size`` and ``num_iterations`` will tighten the confidence interval.
confidence_level: float, default = 0.99
This argument is only considered when ``include_confidence_band=True``, and can be used to specify the confidence level used
for constructing confidence intervals. For example, if ``confidence_level`` is set to 0.99, then the returned DataFrame will include
columns ``p99_high`` and ``p99_low`` which indicates that the true feature importance will be between ``p99_high`` and ``p99_low`` 99% of
the time (99% confidence interval). More generally, if ``confidence_level`` = 0.XX, then the columns containing the XX% confidence interval
will be named ``pXX_high`` and ``pXX_low``.
Returns
-------
:class:`pd.DataFrame` of feature importance scores with 2 columns:
index: The feature name.
'importance': The estimated feature importance score.
'stddev': The standard deviation of the feature importance score. If NaN, then not enough ``num_iterations`` were used.
"""
if data is not None:
data = self._check_and_prepare_data_frame(data)
self._check_data_for_evaluation(data)
fi_df = self._learner.get_feature_importance(
data=data,
model=model,
metric=metric,
features=features,
time_limit=time_limit,
method=method,
subsample_size=subsample_size,
num_iterations=num_iterations,
random_seed=random_seed,
relative_scores=relative_scores,
include_confidence_band=include_confidence_band,
confidence_level=confidence_level,
)
return fi_df
@classmethod
def _load_version_file(cls, path: str) -> str:
"""
Loads the version file that is part of the saved predictor artifact.
Parameters
----------
path: str
The path that would be used to load the predictor via `predictor.load(path)`
Returns
-------
The version of AutoGluon used to fit the predictor, as a string.
"""
version_file_path = os.path.join(path, cls._predictor_version_file_name)
try:
version = load_str.load(path=version_file_path)
except:
# Loads the old version file used in `autogluon.timeseries<=1.1.0`, named `__version__`.
# This file name was changed because Kaggle does not allow uploading files named `__version__`.
version_file_path = os.path.join(path, "__version__")
version = load_str.load(path=version_file_path)
return version
[docs]
@classmethod
def load(cls, path: Union[str, Path], require_version_match: bool = True) -> "TimeSeriesPredictor":
"""Load an existing ``TimeSeriesPredictor`` from given ``path``.
.. warning::
:meth:`autogluon.timeseries.TimeSeriesPredictor.load` uses `pickle` module implicitly, which is known to
be insecure. It is possible to construct malicious pickle data which will execute arbitrary code during
unpickling. Never load data that could have come from an untrusted source, or that could have been tampered
with. **Only load data you trust.**
Parameters
----------
path : str or pathlib.Path
Path where the predictor was saved via :meth:`~autogluon.timeseries.TimeSeriesPredictor.save`.
require_version_match : bool, default = True
If True, will raise an AssertionError if the ``autogluon.timeseries`` version of the loaded predictor does
not match the installed version of ``autogluon.timeseries``.
If False, will allow loading of models trained on incompatible versions, but is NOT recommended. Users may
run into numerous issues if attempting this.
Returns
-------
predictor : TimeSeriesPredictor
Examples
--------
>>> predictor = TimeSeriesPredictor.load(path_to_predictor)
"""
if not path:
raise ValueError("`path` cannot be None or empty in load().")
path: str = setup_outputdir(path, warn_if_exist=False)
try:
version_saved = cls._load_version_file(path=path)
except:
logger.warning(
f'WARNING: Could not find version file at "{os.path.join(path, cls._predictor_version_file_name)}".\n'
f"This means that the predictor was fit in an AutoGluon version `<=0.7.0`."
)
version_saved = "Unknown (Likely <=0.7.0)"
check_saved_predictor_version(
version_current=current_ag_version,
version_saved=version_saved,
require_version_match=require_version_match,
logger=logger,
)
logger.info(f"Loading predictor from path {path}")
learner = AbstractLearner.load(path)
predictor = load_pkl.load(path=os.path.join(learner.path, cls.predictor_file_name))
predictor._learner = learner
predictor.path = learner.path
return predictor
def _save_version_file(self):
version_file_contents = current_ag_version
version_file_path = os.path.join(self.path, self._predictor_version_file_name)
save_str.save(path=version_file_path, data=version_file_contents, verbose=False)
[docs]
def save(self) -> None:
"""Save this predictor to file in directory specified by this Predictor's ``path``.
Note that :meth:`~autogluon.timeseries.TimeSeriesPredictor.fit` already saves the predictor object automatically
(we do not recommend modifying the Predictor object yourself as it tracks many trained models).
"""
tmp_learner = self._learner
self._learner = None
save_pkl.save(path=os.path.join(tmp_learner.path, self.predictor_file_name), object=self)
self._learner = tmp_learner
self._save_version_file()
[docs]
def info(self) -> Dict[str, Any]:
"""Returns a dictionary of objects each describing an attribute of the training process and trained models."""
return self._learner.get_info(include_model_info=True)
@property
def model_best(self) -> str:
"""Returns the name of the best model from trainer."""
if self._trainer.model_best is not None:
models = self._trainer.get_model_names()
if self._trainer.model_best in models:
return self._trainer.model_best
return self._trainer.get_model_best()
[docs]
def persist(
self, models: Union[Literal["all", "best"], List[str]] = "best", with_ancestors: bool = True
) -> List[str]:
"""Persist models in memory for reduced inference latency. This is particularly important if the models are being used for online
inference where low latency is critical. If models are not persisted in memory, they are loaded from disk every time they are
asked to make predictions. This is especially cumbersome for large deep learning based models which have to be loaded into
accelerator (e.g., GPU) memory each time.
Parameters
----------
models : list of str or str, default = 'best'
Model names of models to persist.
If 'best' then the model with the highest validation score is persisted (this is the model used for prediction by default).
If 'all' then all models are persisted. Valid models are listed in this `predictor` by calling `predictor.model_names()`.
with_ancestors : bool, default = True
If True, all ancestor models of the provided models will also be persisted.
If False, ensemble models will not have the models they depend on persisted unless those models were specified in `models`.
This will slow down inference as the ancestor models will still need to be loaded from disk for each predict call.
Only relevant for ensemble models.
Returns
-------
list_of_models : List[str]
List of persisted model names.
"""
return self._learner.persist_trainer(models=models, with_ancestors=with_ancestors)
[docs]
def unpersist(self) -> List[str]:
"""Unpersist models in memory for reduced memory usage. If models are not persisted in memory, they are loaded from
disk every time they are asked to make predictions.
Note: Another way to reset the predictor and unpersist models is to reload the predictor from disk
via `predictor = TimeSeriesPredictor.load(predictor.path)`.
Returns
-------
list_of_models : List[str]
List of unpersisted model names.
"""
return self._learner.unpersist_trainer()
[docs]
def leaderboard(
self,
data: Optional[Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]] = None,
display: bool = False,
use_cache: bool = True,
**kwargs,
) -> pd.DataFrame:
"""Return a leaderboard showing the performance of every trained model, the output is a
pandas data frame with columns:
* ``model``: The name of the model.
* ``score_test``: The test score of the model on ``data``, if provided. Computed according to ``eval_metric``.
* ``score_val``: The validation score of the model using the internal validation data. Computed according to ``eval_metric``.
.. note::
Metrics scores are always shown in 'higher is better' format.
This means that metrics such as MASE or MAPE will be multiplied by -1, so their values will be negative.
This is necessary to avoid the user needing to know the metric to understand if higher is better when
looking at leaderboard.
* ``pred_time_val``: Time taken by the model to predict on the validation data set
* ``fit_time_marginal``: The fit time required to train the model (ignoring base models for ensembles).
* ``fit_order``: The order in which models were fit. The first model fit has ``fit_order=1``, and the Nth
model fit has ``fit_order=N``.
Parameters
----------
data : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str], optional
dataset used for additional evaluation. Must include both historic and future data (i.e., length of all
time series in ``data`` must be at least ``prediction_length + 1``).
If ``known_covariates_names`` were specified when creating the predictor, ``data`` must include the columns
listed in ``known_covariates_names`` with the covariates values aligned with the target time series.
If ``train_data`` used to train the predictor contained past covariates or static features, then ``data``
must also include them (with same column names and dtypes).
If provided data is a path or a pandas.DataFrame, AutoGluon will attempt to automatically convert it to a
``TimeSeriesDataFrame``.
display : bool, default = False
If True, the leaderboard DataFrame will be printed.
use_cache : bool, default = True
If True, will attempt to use the cached predictions. If False, cached predictions will be ignored.
This argument is ignored if ``cache_predictions`` was set to False when creating the ``TimeSeriesPredictor``.
Returns
-------
leaderboard : pandas.DataFrame
The leaderboard containing information on all models and in order of best model to worst in terms of
test performance.
"""
if "silent" in kwargs:
# keep `silent` logic for backwards compatibility
assert isinstance(kwargs["silent"], bool)
display = not kwargs.pop("silent")
if len(kwargs) > 0:
for key in kwargs:
raise TypeError(f"TimeSeriesPredictor.leaderboard() got an unexpected keyword argument '{key}'")
if data is not None:
data = self._check_and_prepare_data_frame(data)
self._check_data_for_evaluation(data)
leaderboard = self._learner.leaderboard(data, use_cache=use_cache)
if display:
with pd.option_context("display.max_rows", None, "display.max_columns", None, "display.width", 1000):
print(leaderboard)
return leaderboard
[docs]
def fit_summary(self, verbosity: int = 1) -> Dict[str, Any]:
"""Output summary of information about models produced during
:meth:`~autogluon.timeseries.TimeSeriesPredictor.fit`.
Parameters
----------
verbosity : int, default = 1
Controls the detail level of summary to output. Set 0 for no output printing.
Returns
-------
summary_dict : Dict[str, Any]
Dict containing various detailed information. We do not recommend directly printing this dict as it may
be very large.
"""
# TODO: HPO-specific information currently not reported in fit_summary
# TODO: Revisit after ray tune integration
model_types = self._trainer.get_models_attribute_dict(attribute="type")
model_typenames = {key: model_types[key].__name__ for key in model_types}
unique_model_types = set(model_typenames.values()) # no more class info
# all fit() information that is returned:
results = {
"model_types": model_typenames, # dict with key = model-name, value = type of model (class-name)
"model_performance": self._trainer.get_models_attribute_dict("val_score"),
"model_best": self._trainer.get_model_best(), # the name of the best model (on validation data)
"model_paths": self._trainer.get_models_attribute_dict("path"),
"model_fit_times": self._trainer.get_models_attribute_dict("fit_time"),
"model_pred_times": self._trainer.get_models_attribute_dict("predict_time"),
}
# get dict mapping model name to final hyperparameter values for each model:
model_hyperparams = {}
for model_name in self.model_names():
model_obj = self._trainer.load_model(model_name)
model_hyperparams[model_name] = model_obj.params
results["model_hyperparams"] = model_hyperparams
results["leaderboard"] = self._learner.leaderboard()
if verbosity > 0: # print stuff
print("****************** Summary of fit() ******************")
print("Estimated performance of each model:")
print(results["leaderboard"])
print(f"Number of models trained: {len(results['model_performance'])}")
print("Types of models trained:")
print(unique_model_types)
print("****************** End of fit() summary ******************")
return results
[docs]
def refit_full(self, model: str = "all", set_best_to_refit_full: bool = True) -> Dict[str, str]:
"""Retrain model on all of the data (training + validation).
This method can only be used if no ``tuning_data`` was passed to :meth:`~autogluon.timeseries.TimeSeriesPredictor.fit`.
.. warning::
This is experimental functionality, many time series models do not yet support ``refit_full`` and will
simply be copied.
Parameters
----------
model : str, default = "all"
Name of the model to refit.
All ancestor models will also be refit in the case that the selected model is a weighted ensemble.
Valid models are listed in this ``predictor`` by calling :meth:`~autogluon.timeseries.TimeSeriesPredictor.model_names`.
* If "all" then all models are refitted.
* If "best" then the model with the highest validation score is refit.
set_best_to_refit_full : bool, default = True
If True, sets best model to the refit_full version of the prior best model. This means the model used when
``predictor.predict(data)`` is called will be the refit_full version instead of the original version of the
model. Has no effect if ``model`` is not the best model.
"""
logger.warning(
"\tWARNING: refit_full functionality for TimeSeriesPredictor is experimental "
"and is not yet supported by all models."
)
logger.info(
"Refitting models via `refit_full` using all of the data (combined train and validation)...\n"
"\tModels trained in this way will have the suffix '_FULL' and have NaN validation score.\n"
"\tThis process is not bound by time_limit, but should take less time than the original `fit` call."
)
model_best = self.model_best
refit_full_dict = self._learner.refit_full(model=model)
if set_best_to_refit_full:
if model_best in refit_full_dict:
self._trainer.model_best = refit_full_dict[model_best]
self._trainer.save()
logger.info(
f"Updated best model to '{self._trainer.model_best}' (Previously '{model_best}'). "
f"AutoGluon will default to using '{self._trainer.model_best}' for predict()."
)
elif model_best in refit_full_dict.values():
# Model best is already a refit full model
prev_best = self._trainer.model_best
self._trainer.model_best = model_best
self._trainer.save()
logger.info(
f"Updated best model to '{self._trainer.model_best}' (Previously '{prev_best}'). "
f"AutoGluon will default to using '{self._trainer.model_best}' for predict()."
)
else:
logger.warning(
f"Best model ('{model_best}') is not present in refit_full dictionary. "
f"Training may have failed on the refit model. AutoGluon will default to using '{model_best}' for predict()."
)
return refit_full_dict
def __dir__(self) -> List[str]:
# This hides method from IPython autocomplete, but not VSCode autocomplete
deprecated = ["score", "get_model_best", "get_model_names"]
return [d for d in super().__dir__() if d not in deprecated]
def _simulation_artifact(self, test_data: TimeSeriesDataFrame) -> dict:
"""[Advanced] Computes and returns the necessary information to perform offline ensemble simulation."""
def select_target(ts_df: TimeSeriesDataFrame) -> TimeSeriesDataFrame:
ts_df = ts_df.copy()
ts_df.static_features = None
return ts_df[[self.target]]
test_data = self._check_and_prepare_data_frame(test_data)
self._check_data_for_evaluation(test_data, name="test_data")
test_data = self._learner.feature_generator.transform(test_data)
trainer = self._trainer
train_data = trainer.load_train_data()
val_data = trainer.load_val_data()
base_models = trainer.get_model_names(level=0)
pred_proba_dict_val: Dict[str, List[TimeSeriesDataFrame]] = {
model: trainer._get_model_oof_predictions(model) for model in base_models
}
past_data, known_covariates = test_data.get_model_inputs_for_scoring(
prediction_length=self.prediction_length, known_covariates_names=trainer.metadata.known_covariates
)
pred_proba_dict_test: Dict[str, TimeSeriesDataFrame] = trainer.get_model_pred_dict(
base_models, data=past_data, known_covariates=known_covariates
)
y_val: List[TimeSeriesDataFrame] = [
select_target(df) for df in trainer._get_ensemble_oof_data(train_data=train_data, val_data=val_data)
]
y_test: TimeSeriesDataFrame = select_target(test_data)
simulation_dict = dict(
pred_proba_dict_val=pred_proba_dict_val,
pred_proba_dict_test=pred_proba_dict_test,
y_val=y_val,
y_test=y_test,
target=self.target,
prediction_length=self.prediction_length,
eval_metric=self.eval_metric.name,
eval_metric_seasonal_period=self.eval_metric_seasonal_period,
quantile_levels=self.quantile_levels,
)
return simulation_dict
[docs]
def plot(
self,
data: Union[TimeSeriesDataFrame, pd.DataFrame, Path, str],
predictions: Optional[TimeSeriesDataFrame] = None,
quantile_levels: Optional[List[float]] = None,
item_ids: Optional[List[Union[str, int]]] = None,
max_num_item_ids: int = 8,
max_history_length: Optional[int] = None,
point_forecast_column: Optional[str] = None,
matplotlib_rc_params: Optional[dict] = None,
):
"""Plot historic time series values and the forecasts.
Parameters
----------
data : Union[TimeSeriesDataFrame, pd.DataFrame, Path, str]
Observed time series data.
predictions : TimeSeriesDataFrame, optional
Predictions generated by calling :meth:`~autogluon.timeseries.TimeSeriesPredictor.predict`.
quantile_levels : List[float], optional
Quantile levels for which to plot the prediction intervals. Defaults to lowest & highest quantile levels
available in ``predictions``.
item_ids : List[Union[str, int]], optional
If provided, plots will only be generated for time series with these item IDs. By default (if set to
``None``), item IDs are selected randomly. In either case, plots are generated for at most
``max_num_item_ids`` time series.
max_num_item_ids : int, default = 8
At most this many time series will be plotted by the method.
max_history_length : int, optional
If provided, at most this many time steps will be shown for each time series in ``data``.
point_forecast_column : str, optional
Name of the column in ``predictions`` that will be plotted as the point forecast. Defaults to ``"0.5"``,
if this column is present in ``predictions``, otherwise ``"mean"``.
matplotlib_rc_params : dict, optional
Dictionary describing the plot style that will be passed to [`matplotlib.pyplot.rc_context`](https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.rc_context.html).
See [matplotlib documentation](https://matplotlib.org/stable/users/explain/customizing.html#the-default-matplotlibrc-file) for the list of available options.
"""
import matplotlib.pyplot as plt
data = self._check_and_prepare_data_frame(data)
if item_ids is None:
item_ids = list(np.random.choice(data.item_ids, size=min(max_num_item_ids, data.num_items), replace=False))
else:
item_ids = list(item_ids)[:max_num_item_ids]
if predictions is not None:
if (
not isinstance(predictions, TimeSeriesDataFrame)
or "mean" not in predictions.columns
or predictions.index.nlevels != 2
):
raise ValueError("predictions must be a TimeSeriesDataFrame produced by predictor.predict()")
if point_forecast_column is None:
point_forecast_column = "0.5" if "0.5" in predictions.columns else "mean"
if quantile_levels is None:
available_quantile_levels = [float(q) for q in predictions.columns if q != "mean"]
if len(available_quantile_levels) >= 2:
quantile_levels = [min(available_quantile_levels), max(available_quantile_levels)]
else:
quantile_levels = []
if len(item_ids) == 1:
ncols = 1
nrows = 1
else:
ncols = 2
nrows = math.ceil(len(item_ids) / ncols)
rc_params = {
"font.size": 10,
"figure.figsize": [20, 3.5 * nrows],
"figure.dpi": 100,
"legend.loc": "upper center",
}
if matplotlib_rc_params is not None:
rc_params.update(matplotlib_rc_params)
with plt.rc_context(rc_params):
fig, axes = plt.subplots(ncols=ncols, nrows=nrows, squeeze=False)
fig.tight_layout(h_pad=2.5, w_pad=0.5)
axes = axes.ravel()
for i, (item_id, ax) in enumerate(zip(item_ids, axes)):
ax.set_title(item_id)
ax.grid()
# Label the x axis for subplots in the lowest row
if i // nrows == 1:
ax.set_xlabel("Time")
# Label the y axis for subplots in the leftmost column
if i % ncols == 0:
ax.set_ylabel(self.target)
ts = data.loc[item_id][self.target]
if max_history_length is not None:
ts = ts.iloc[-max_history_length:]
ax.plot(ts, label="Observed", color="C0")
if predictions is not None:
forecast = predictions.loc[item_id]
point_forecast = forecast[point_forecast_column]
ax.plot(point_forecast, color="C1", label="Forecast")
if quantile_levels is not None:
for q in quantile_levels:
ax.fill_between(forecast.index, point_forecast, forecast[str(q)], color="C1", alpha=0.2)
if len(axes) > len(item_ids):
axes[len(item_ids)].set_axis_off()
handles, labels = axes[0].get_legend_handles_labels()
fig.legend(handles, labels, bbox_to_anchor=(0.5, 0.0), ncols=len(handles))
return fig