Datasets

skforecast.datasets.fetch_dataset

fetch_dataset(
    name,
    version="latest",
    raw=False,
    kwargs_read_csv={},
    verbose=True,
)

Fetch a dataset from the skforecast-datasets repository.

Parameters:

Name	Type	Description	Default
name	str	Name of the dataset to fetch.	required
version	str | int	Version of the dataset to fetch. If 'latest', the latest version will be fetched (the one in the main branch). For a list of available versions, see the repository branches.	'latest'
raw	bool	If True, the raw dataset is fetched. If False, the preprocessed dataset is fetched. The preprocessing consists of setting the column with the date/time as index and converting the index to datetime. A frequency is also set to the index.	False
kwargs_read_csv	dict	Kwargs to pass to pandas read_csv function.	{}
verbose	bool	If True, print information about the dataset.	True

Returns:

Name	Type	Description
df	pandas DataFrame	Dataset.

Source code in skforecast/datasets/datasets.py

def fetch_dataset(
    name: str,
    version: str | int = 'latest',
    raw: bool = False,
    kwargs_read_csv: dict = {},
    verbose: bool = True
) -> pd.DataFrame:
    """
    Fetch a dataset from the skforecast-datasets repository.

    Parameters
    ----------
    name: str
        Name of the dataset to fetch.
    version: str, int, default `'latest'`
        Version of the dataset to fetch. If 'latest', the latest version will be 
        fetched (the one in the main branch). For a list of available versions, 
        see the repository branches.
    raw: bool, default `False`
        If True, the raw dataset is fetched. If False, the preprocessed dataset 
        is fetched. The preprocessing consists of setting the column with the 
        date/time as index and converting the index to datetime. A frequency is 
        also set to the index.
    kwargs_read_csv: dict, default `{}`
        Kwargs to pass to pandas `read_csv` function.
    verbose: bool, default `True`
        If True, print information about the dataset.

    Returns
    -------
    df: pandas DataFrame
        Dataset.

    """

    version = 'main' if version == 'latest' else f'{version}'

    if name not in datasets.keys():
        raise ValueError(
            f"Dataset '{name}' not found. "
            f"Available datasets are: {sorted(datasets.keys())}"
        )

    url = datasets[name]['url']
    if '{version}' in url:
        url = url.format(version=version)
    file_type = datasets[name]['file_type']

    if not isinstance(url, list):
        parsed = urlparse(url)
        if parsed.scheme == "https" and parsed.netloc == "drive.google.com":
            file_id = url.split('/')[-2]
            url = 'https://drive.google.com/uc?id=' + file_id
        if file_type == 'csv':
            try:
                sep = datasets[name]['sep']
                df = pd.read_csv(url, sep=sep, **kwargs_read_csv)
            except Exception as e:
                raise ValueError(
                    f"Error reading dataset '{name}' from {url}: {str(e)}."
                )
        if file_type == 'parquet':
            try:
                df = pd.read_parquet(url)
            except Exception as e:
                raise ValueError(
                    f"Error reading dataset '{name}' from {url}: {str(e)}."
                )
    else:
        try: 
            df = []
            for url_partition in url:
                path = 'https://drive.google.com/uc?export=download&id=' + url_partition.split('/')[-2]
                df.append(pd.read_parquet(path))
        except Exception as e:
            raise ValueError(
                f"Error reading dataset '{name}' from {url}: {str(e)}."
            )
        df = pd.concat(df, axis=0).reset_index(drop=True)

    if not raw:
        try:
            index_col = datasets[name]['index_col']
            freq = datasets[name]['freq']
            if freq == 'H' and pd.__version__ >= '2.2.0':
                freq = "h"
            date_format = datasets[name]['date_format']
            df = df.set_index(index_col)
            df.index = pd.to_datetime(df.index, format=date_format)
            df = df.asfreq(freq)
            df = df.sort_index()
        except:
            pass

    if verbose:
        _print_dataset_info(name, version=version, shape=df.shape)

    return df

skforecast.datasets.load_demo_dataset

load_demo_dataset(version='latest')

Load demo data set with monthly expenditure ($AUD) on corticosteroid drugs that the Australian health system had between 1991 and 2008. Obtained from the book: Forecasting: Principles and Practice by Rob J Hyndman and George Athanasopoulos. Index is set to datetime with monthly frequency and sorted.

Parameters:

Name	Type	Description	Default
version	str	Version of the dataset to fetch. If 'latest', the latest version will be fetched (the one in the main branch). For a list of available versions, see the repository branches.	'latest'

Returns:

Name	Type	Description
df	pandas Series	Dataset.

Source code in skforecast/datasets/datasets.py

def load_demo_dataset(version: str = 'latest') -> pd.Series:
    """
    Load demo data set with monthly expenditure ($AUD) on corticosteroid drugs that
    the Australian health system had between 1991 and 2008. Obtained from the book:
    Forecasting: Principles and Practice by Rob J Hyndman and George Athanasopoulos.
    Index is set to datetime with monthly frequency and sorted.

    Parameters
    ----------
    version: str, default `'latest'`
        Version of the dataset to fetch. If 'latest', the latest version will be
        fetched (the one in the main branch). For a list of available versions,
        see the repository branches.

    Returns
    -------
    df: pandas Series
        Dataset.

    """

    version = 'main' if version == 'latest' else f'{version}'

    url = (
        f'https://raw.githubusercontent.com/skforecast/skforecast-datasets/{version}/'
        'data/h2o.csv'
    )

    df = pd.read_csv(url, sep=',', header=0, names=['y', 'datetime'])
    df['datetime'] = pd.to_datetime(df['datetime'], format='%Y-%m-%d')
    df = df.set_index('datetime')
    df = df.asfreq('MS')
    df = df['y']
    df = df.sort_index()

    return df

skforecast.datasets.show_datasets_info

show_datasets_info(datasets_names=None, version='latest')

Print information about available datasets. If datasets_names is provided, only information about those datasets will be printed.

Parameters:

Name	Type	Description	Default
datasets_names	list[str] | None	List of dataset names to display information about. If None, information about all datasets will be displayed.	None
version	str	Version of the datasets to display information about.	'latest'

Returns:

Type	Description
None

Source code in skforecast/datasets/datasets.py

def show_datasets_info(
    datasets_names: list[str] | None = None,
    version: str = 'latest'
) -> None:
    """
    Print information about available datasets. If `datasets_names` is provided,
    only information about those datasets will be printed.

    Parameters
    ----------
    datasets_names: list, default None
        List of dataset names to display information about. If None, information 
        about all datasets will be displayed.
    version: str
        Version of the datasets to display information about.

    Returns
    -------
    None

    """

    datasets_names = datasets_names or sorted(datasets.keys())
    version = 'main' if version == 'latest' else f'{version}'

    for dataset_name in datasets_names:
        if dataset_name in datasets:
            _print_dataset_info(dataset_name, version=version)
        else:
            print(
                f"Dataset '{dataset_name}' not available. Set argument "
                f"`datasets_names` to None to see all available datasets."
            )