Datasetsยถ
All datasets used in the skforecast library and related tutorials are accessible using the skforecast.datasets.fetch_dataset() function.
Each dataset in this collection comes with a description of its time series and a reference to its original source.
Available data sets are stored at skforecast-datasets and can be easily listed using the skforecast.datasets.show_datasets_info() function.
In [1]:
Copied!
# Libraries
# ==============================================================================
from skforecast.datasets import fetch_dataset, show_datasets_info
# Libraries
# ==============================================================================
from skforecast.datasets import fetch_dataset, show_datasets_info
By default, the data is structured as a pandas dataframe with a datetime index and frequency. Additionally, a concise description is printed for quick reference.
In [2]:
Copied!
# Download data
# ==============================================================================
data = fetch_dataset(name="bike_sharing")
data.head()
# Download data
# ==============================================================================
data = fetch_dataset(name="bike_sharing")
data.head()
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bike_sharing โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly usage of the bike share system in the city of Washington D.C. during the years 2011 and 2012. In โ โ addition to the number of users per hour, information about weather conditions and holidays is available. โ โ โ โ Source: โ โ Fanaee-T,Hadi. (2013). Bike Sharing Dataset. UCI Machine Learning Repository. https://doi.org/10.24432/C5W894. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/bike_sharing_dataset_clean.csv โ โ โ โ Shape: 17544 rows x 11 columns โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
Out[2]:
| holiday | workingday | weather | temp | atemp | hum | windspeed | users | month | hour | weekday | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| date_time | |||||||||||
| 2011-01-01 00:00:00 | 0.0 | 0.0 | clear | 9.84 | 14.395 | 81.0 | 0.0 | 16.0 | 1 | 0 | 5 |
| 2011-01-01 01:00:00 | 0.0 | 0.0 | clear | 9.02 | 13.635 | 80.0 | 0.0 | 40.0 | 1 | 1 | 5 |
| 2011-01-01 02:00:00 | 0.0 | 0.0 | clear | 9.02 | 13.635 | 80.0 | 0.0 | 32.0 | 1 | 2 | 5 |
| 2011-01-01 03:00:00 | 0.0 | 0.0 | clear | 9.84 | 14.395 | 75.0 | 0.0 | 13.0 | 1 | 3 | 5 |
| 2011-01-01 04:00:00 | 0.0 | 0.0 | clear | 9.84 | 14.395 | 75.0 | 0.0 | 1.0 | 1 | 4 | 5 |
Downloading raw data, without any preprocessing, is possible by specifying the raw=True argument.
In [3]:
Copied!
# Download raw data
# ==============================================================================
data = fetch_dataset(name="bike_sharing", raw=True)
data.head()
# Download raw data
# ==============================================================================
data = fetch_dataset(name="bike_sharing", raw=True)
data.head()
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bike_sharing โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly usage of the bike share system in the city of Washington D.C. during the years 2011 and 2012. In โ โ addition to the number of users per hour, information about weather conditions and holidays is available. โ โ โ โ Source: โ โ Fanaee-T,Hadi. (2013). Bike Sharing Dataset. UCI Machine Learning Repository. https://doi.org/10.24432/C5W894. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/bike_sharing_dataset_clean.csv โ โ โ โ Shape: 17544 rows x 12 columns โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
Out[3]:
| date_time | holiday | workingday | weather | temp | atemp | hum | windspeed | users | month | hour | weekday | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 2011-01-01 00:00:00 | 0.0 | 0.0 | clear | 9.84 | 14.395 | 81.0 | 0.0 | 16.0 | 1 | 0 | 5 |
| 1 | 2011-01-01 01:00:00 | 0.0 | 0.0 | clear | 9.02 | 13.635 | 80.0 | 0.0 | 40.0 | 1 | 1 | 5 |
| 2 | 2011-01-01 02:00:00 | 0.0 | 0.0 | clear | 9.02 | 13.635 | 80.0 | 0.0 | 32.0 | 1 | 2 | 5 |
| 3 | 2011-01-01 03:00:00 | 0.0 | 0.0 | clear | 9.84 | 14.395 | 75.0 | 0.0 | 13.0 | 1 | 3 | 5 |
| 4 | 2011-01-01 04:00:00 | 0.0 | 0.0 | clear | 9.84 | 14.395 | 75.0 | 0.0 | 1.0 | 1 | 4 | 5 |
Datasets with several series in long format (one row per series and date, see their description) have repeated dates in the index. For them, no frequency is set and the rows are sorted by date, keeping the order of the series within each date. Use reshape_series_long_to_dict (after reset_index, it needs the dates as a column) to get one series per key.
When the dates of a dataset are not complete, setting the frequency introduces rows with NaN values for the missing dates, and fetch_dataset warns with a MissingValuesWarning.
In [4]:
Copied!
# Show dataset information
# ==============================================================================
show_datasets_info(datasets_names=None) # None means all datasets
# Show dataset information
# ==============================================================================
show_datasets_info(datasets_names=None) # None means all datasets
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ air_quality_valencia โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly measures of several air chemical pollutants at Valencia city (Avd. Francia) from 2019-01-01 to โ โ 2023-12-31. Including the following variables: pm2.5 (ยตg/mยณ), CO (mg/mยณ), NO (ยตg/mยณ), NO2 (ยตg/mยณ), PM10 โ โ (ยตg/mยณ), NOx (ยตg/mยณ), O3 (ยตg/mยณ), Veloc. (m/s), Direc. (degrees), SO2 (ยตg/mยณ). โ โ โ โ Source: โ โ Red de Vigilancia y Control de la Contaminaciรณn Atmosfรฉrica, 46250047-Valรจncia - Av. Franรงa, โ โ https://mediambient.gva.es/es/web/calidad-ambiental/datos-historicos. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/air_quality_valencia.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ air_quality_valencia_no_missing โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly measures of several air chemical pollutants at Valencia city (Avd. Francia) from 2019-01-01 to โ โ 2023-12-31. Including the following variables: pm2.5 (ยตg/mยณ), CO (mg/mยณ), NO (ยตg/mยณ), NO2 (ยตg/mยณ), PM10 โ โ (ยตg/mยณ), NOx (ยตg/mยณ), O3 (ยตg/mยณ), Veloc. (m/s), Direc. (degrees), SO2 (ยตg/mยณ). Missing values have been imputed โ โ using linear interpolation. โ โ โ โ Source: โ โ Red de Vigilancia y Control de la Contaminaciรณn Atmosfรฉrica, 46250047-Valรจncia - Av. Franรงa, โ โ https://mediambient.gva.es/es/web/calidad-ambiental/datos-historicos. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/air_quality_valencia_no_missing.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ ashrae_daily โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Daily energy consumption data from the ASHRAE competition with building metadata and weather data. Long format: โ โ one row per building and date. โ โ โ โ Source: โ โ Kaggle competition Addison Howard, Chris Balbach, Clayton Miller, Jeff Haberl, Krishnan Gowri, Sohier Dane. โ โ (2019). ASHRAE - Great Energy Predictor III. Kaggle. https://www.kaggle.com/c/ashrae-energy-prediction/overview โ โ โ โ URL: โ โ https://huggingface.co/datasets/skforecast/ashrae_daily/resolve/main/ashrae_daily.parquet โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ australia_tourism โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Quarterly overnight trips (in thousands) from 1998 Q1 to 2017 Q4 across Australia. The tourism regions are โ โ formed through the aggregation of Statistical Local Areas (SLAs) which are defined by the various State and โ โ Territory tourism authorities according to their research and marketing needs. Long format: one row per region, โ โ state, purpose and quarter (304 series). โ โ โ โ Source: โ โ Wang, E, D Cook, and RJ Hyndman (2020). A new tidy data structure to support exploration and modeling of โ โ temporal data, Journal of Computational and Graphical Statistics, 29:3, 466-478, โ โ doi:10.1080/10618600.2019.1695624. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/australia_tourism.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bdg2_daily โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Daily energy consumption data from the Building Data Genome Project 2 with building metadata and weather data. โ โ Long format: one row per building and date. https://github.com/buds-lab/building-data-genome-project-2 โ โ โ โ Source: โ โ Miller, C., Kathirgamanathan, A., Picchetti, B. et al. The Building Data Genome Project 2, energy meter data โ โ from the ASHRAE Great Energy Predictor III competition. Sci Data 7, 368 (2020). โ โ https://doi.org/10.1038/s41597-020-00712-x โ โ โ โ URL: โ โ https://huggingface.co/datasets/skforecast/bdg2_daily/resolve/main/bdg2_daily.parquet โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bdg2_daily_sample โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Daily energy consumption data of two buildings sampled from the Building Data Genome Project 2. โ โ https://github.com/buds-lab/building-data-genome-project-2 โ โ โ โ Source: โ โ Miller, C., Kathirgamanathan, A., Picchetti, B. et al. The Building Data Genome Project 2, energy meter data โ โ from the ASHRAE Great Energy Predictor III competition. Sci Data 7, 368 (2020). โ โ https://doi.org/10.1038/s41597-020-00712-x โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/bdg2_daily_sample.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bdg2_hourly โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly energy consumption data from the Building Data Genome Project 2 with building metadata and weather data. โ โ Long format: one row per building and timestamp. https://github.com/buds-lab/building-data-genome-project-2 โ โ โ โ Source: โ โ Miller, C., Kathirgamanathan, A., Picchetti, B. et al. The Building Data Genome Project 2, energy meter data โ โ from the ASHRAE Great Energy Predictor III competition. Sci Data 7, 368 (2020). โ โ https://doi.org/10.1038/s41597-020-00712-x โ โ โ โ URL: โ โ https://huggingface.co/datasets/skforecast/bdg2_hourly/resolve/main/bdg2_hourly.parquet โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bdg2_hourly_sample โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly energy consumption data of two buildings sampled from the Building Data Genome Project 2. โ โ https://github.com/buds-lab/building-data-genome-project-2 โ โ โ โ Source: โ โ Miller, C., Kathirgamanathan, A., Picchetti, B. et al. The Building Data Genome Project 2, energy meter data โ โ from the ASHRAE Great Energy Predictor III competition. Sci Data 7, 368 (2020). โ โ https://doi.org/10.1038/s41597-020-00712-x โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/bdg2_hourly_sample.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bicimad โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ This dataset contains the daily users of the bicycle rental service (BiciMad) in the city of Madrid (Spain) โ โ from 2014-06-23 to 2022-09-30. โ โ โ โ Source: โ โ The original data was obtained from: Portal de datos abiertos del Ayuntamiento de Madrid โ โ https://datos.madrid.es/portal/site/egob โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/bicimad_users.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bike_sharing โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly usage of the bike share system in the city of Washington D.C. during the years 2011 and 2012. In โ โ addition to the number of users per hour, information about weather conditions and holidays is available. โ โ โ โ Source: โ โ Fanaee-T,Hadi. (2013). Bike Sharing Dataset. UCI Machine Learning Repository. https://doi.org/10.24432/C5W894. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/bike_sharing_dataset_clean.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ bike_sharing_extended_features โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly usage of the bike share system in the city of Washington D.C. during the years 2011 and 2012. In โ โ addition to the number of users per hour, the dataset was enriched by introducing supplementary features. โ โ Addition includes calendar-based variables (day of the week, hour of the day, month, etc.), indicators for โ โ sunlight, incorporation of rolling temperature averages, and the creation of polynomial features generated from โ โ variable pairs. All cyclic variables are encoded using sine and cosine functions to ensure accurate โ โ representation. โ โ โ โ Source: โ โ Fanaee-T,Hadi. (2013). Bike Sharing Dataset. UCI Machine Learning Repository. https://doi.org/10.24432/C5W894. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/bike_sharing_extended_features.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ ett_m1 โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Data from an electricity transformer station was collected from 2016-07-01 to 2018-06-26 (69,680 data points โ โ every 15 minutes). Each data point consists of 8 features, including the date of the point, the predictive โ โ value "Oil Temperature (OT)", and 6 different types of external power load features: High UseFul Load (HUFL), โ โ High UseLess Load (HULL), Middle UseFul Load (MUFL), Middle UseLess Load (MULL), Low UseFul Load (LUFL), Low โ โ UseLess Load (LULL). โ โ โ โ Source: โ โ Zhou, Haoyi & Zhang, Shanghang & Peng, Jieqi & Zhang, Shuai & Li, Jianxin & Xiong, Hui & Zhang, Wancai. (2020). โ โ Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. โ โ [10.48550/arXiv.2012.07436](https://arxiv.org/abs/2012.07436). https://github.com/zhouhaoyi/ETDataset โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/ETTm1.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ ett_m2 โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Data from an electricity transformer station was collected from 2016-07-01 to 2018-06-26 (69,680 data points โ โ every 15 minutes). Each data point consists of 8 features, including the date of the point, the predictive โ โ value "Oil Temperature (OT)", and 6 different types of external power load features: High UseFul Load (HUFL), โ โ High UseLess Load (HULL), Middle UseFul Load (MUFL), Middle UseLess Load (MULL), Low UseFul Load (LUFL), Low โ โ UseLess Load (LULL). โ โ โ โ Source: โ โ Zhou, Haoyi & Zhang, Shanghang & Peng, Jieqi & Zhang, Shuai & Li, Jianxin & Xiong, Hui & Zhang, Wancai. (2020). โ โ Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. โ โ [10.48550/arXiv.2012.07436](https://arxiv.org/abs/2012.07436). https://github.com/zhouhaoyi/ETDataset โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/ETTm2.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ ett_m2_extended โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Data from an electricity transformer station was collected from 2016-07-01 to 2018-06-26 (69,680 data points โ โ every 15 minutes). Each data point consists of 8 features, including the date of the point, the predictive โ โ value "Oil Temperature (OT)", and 6 different types of external power load features: High UseFul Load (HUFL), โ โ High UseLess Load (HULL), Middle UseFul Load (MUFL), Middle UseLess Load (MULL), Low UseFul Load (LUFL), Low โ โ UseLess Load (LULL). Additional variables are created based on calendar information (year, month, week, day of โ โ the week, and hour). These variables have been encoded using the cyclical encoding technique (sin and cos โ โ transformations) to preserve the cyclical nature of the data. โ โ โ โ Source: โ โ Zhou, Haoyi & Zhang, Shanghang & Peng, Jieqi & Zhang, Shuai & Li, Jianxin & Xiong, Hui & Zhang, Wancai. (2020). โ โ Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. โ โ [10.48550/arXiv.2012.07436](https://arxiv.org/abs/2012.07436). https://github.com/zhouhaoyi/ETDataset โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/ETTm2_extended.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ expenditures_australia โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Monthly expenditure on cafes, restaurants and takeaway food services in Victoria (Australia) from April 1982 up โ โ to April 2024. โ โ โ โ Source: โ โ Australian Bureau of Statistics. Catalogue No. 8501.0 โ โ https://www.abs.gov.au/statistics/industry/retail-and-wholesale-trade/retail-trade-australia/apr-2024/8501011.x โ โ lsx โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/expenditures_australia.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ fuel_consumption โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Monthly fuel consumption in Spain from 1969-01-01 to 2022-08-01. โ โ โ โ Source: โ โ Obtained from Corporaciรณn de Reservas Estratรฉgicas de Productos Petrolรญferos and Corporaciรณn de Derecho Pรบblico โ โ tutelada por el Ministerio para la Transiciรณn Ecolรณgica y el Reto Demogrรกfico. โ โ https://www.cores.es/es/estadisticas โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/consumos-combustibles-mensual.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ h2o โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Monthly expenditure ($AUD) on corticosteroid drugs that the Australian health system had between 1991 and 2008. โ โ โ โ Source: โ โ Hyndman R (2023). fpp3: Data for Forecasting: Principles and Practice (3rd Edition). โ โ http://pkg.robjhyndman.com/fpp3package/, https://github.com/robjhyndman/fpp3package, http://OTexts.com/fpp3. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/h2o.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ h2o_exog โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Monthly expenditure ($AUD) on corticosteroid drugs that the Australian health system had between 1991 and 2008. โ โ Two additional variables (exog_1, exog_2) are simulated. โ โ โ โ Source: โ โ Hyndman R (2023). fpp3: Data for Forecasting: Principles and Practice (3rd Edition). โ โ http://pkg.robjhyndman.com/fpp3package/, https://github.com/robjhyndman/fpp3package, http://OTexts.com/fpp3. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/h2o_exog.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ items_sales โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Simulated time series for the sales of 3 different items. โ โ โ โ Source: โ โ Simulated data. โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/simulated_items_sales.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ m4_daily โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Time series with daily frequency from the M4 competition. Timestamps are artificial (the M4 series have no real โ โ dates): all at 12:00 and spanning 1959 to 2039. Long format: one row per series and timestamp. โ โ โ โ Source: โ โ Monash Time Series Forecasting Repository (https://zenodo.org/communities/forecasting) Godahewa, R., Bergmeir, โ โ C., Webb, G. I., Hyndman, R. J., & Montero-Manso, P. (2021). Monash Time Series Forecasting Archive. In Neural โ โ Information Processing Systems Track on Datasets and Benchmarks. โ โ Raw data, available in .tsf format, has been converted to Pandas format using the code provided by the authors โ โ in https://github.com/rakshitha123/TSForecasting/blob/master/utils/data_loader.py โ โ The category of each time series has been included in the dataset. This information has been obtained from the โ โ Kaggle competition page: https://www.kaggle.com/datasets/yogesh94/m4-forecasting-competition-dataset โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/m4_daily.parquet โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ m4_hourly โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Time series with hourly frequency from the M4 competition. Timestamps are artificial (the M4 series have no โ โ real dates): every series starts at 12:00, and they span 2009 to 2017. Long format: one row per series and โ โ timestamp. โ โ โ โ Source: โ โ Monash Time Series Forecasting Repository (https://zenodo.org/communities/forecasting) Godahewa, R., Bergmeir, โ โ C., Webb, G. I., Hyndman, R. J., & Montero-Manso, P. (2021). Monash Time Series Forecasting Archive. In Neural โ โ Information Processing Systems Track on Datasets and Benchmarks. โ โ Raw data, available in .tsf format, has been converted to Pandas format using the code provided by the authors โ โ in https://github.com/rakshitha123/TSForecasting/blob/master/utils/data_loader.py โ โ The category of each time series has been included in the dataset. This information has been obtained from the โ โ Kaggle competition page: https://www.kaggle.com/datasets/yogesh94/m4-forecasting-competition-dataset โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/m4_hourly.parquet โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ m5 โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Daily sales data from the M5 competition with product metadata and calendar data. Series ids keep the โ โ `_evaluation` suffix of the Kaggle files; `sell_price` is NaN before the product was first sold. Long format: โ โ one row per series and date. โ โ โ โ Source: โ โ Addison Howard, inversion, Spyros Makridakis, and vangelis. M5 Forecasting - Accuracy. โ โ https://kaggle.com/competitions/m5-forecasting-accuracy, 2020. Kaggle. โ โ โ โ URL: โ โ https://huggingface.co/datasets/skforecast/m5_daily/resolve/main/m5_daily.parquet โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ public_transport_madrid โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Daily users of public transport in Madrid (Spain) from 2023-01-01 to 2024-12-15. โ โ โ โ Source: โ โ Consorcio Regional de Transportes de Madrid CRTM, CRTM Evolucion demanda diaria โ โ https://datos.crtm.es/documents/a7210254c4514a19a51b1617cfd61f75/about โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/public-transport-madrid.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ store_sales โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ This dataset contains 913,000 sales transactions from 2013-01-01 to 2017-12-31 for 50 products (SKU) in 10 โ โ stores. Long format: one row per store, item and date. โ โ โ โ Source: โ โ The original data was obtained from: inversion. (2018). Store Item Demand Forecasting Challenge. Kaggle. โ โ https://kaggle.com/competitions/demand-forecasting-kernels-only โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/store_sales.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ turbine_emission โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ The dataset contains 36733 instances of 11 sensor measures aggregated over one hour, from a gas turbine located โ โ in Turkey for the purpose of studying flue gas emissions, namely CO and NOx. Available variables include: โ โ Ambient temperature (AT), Ambient pressure (AP), Ambient humidity (AH), Air filter difference pressure (AFDP), โ โ Gas turbine exhaust pressure (GTEP), Turbine inlet temperature (TIT), Turbine after temperature (TAT), โ โ Compressor discharge pressure (CDP), Turbine energy yield (TEY), Carbon monoxide (CO), and Nitrogen oxides โ โ (NOx). Covered period from 2011-01-01 00:00:00 to 2015-03-11 12:00:00. โ โ โ โ Source: โ โ https://archive.ics.uci.edu/dataset/551/gas+turbine+co+and+nox+emission+data+set โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/turbine_emission.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ uk_daily_flights โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Daily number of flights in UK from 02/01/2019 to 23/01/2022. โ โ โ โ Source: โ โ Experimental statistics published as part of the Economic activity and social change in the UK, real-time โ โ indicators release, Published 27 January 2022. Daily flight numbers are available in the dashboard provided by โ โ the European Organisation for the Safety of Air Navigation (EUROCONTROL). โ โ https://www.ons.gov.uk/economy/economicoutputandproductivity/output/bulletins/economicactivityandsocialchangein โ โ theukrealtimeindicators/latest โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/uk_daily_flights.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ vic_electricity โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Half-hourly electricity demand for Victoria, Australia. `Time` is in UTC (suffix Z, parsed as a naive โ โ datetime); `Date` and `Holiday` refer to local time in Melbourne (UTC+10, UTC+11 in summer). โ โ โ โ Source: โ โ O'Hara-Wild M, Hyndman R, Wang E, Godahewa R (2022). tsibbledata: Diverse Datasets for 'tsibble'. โ โ https://tsibbledata.tidyverts.org/, https://github.com/tidyverts/tsibbledata/. โ โ https://tsibbledata.tidyverts.org/reference/vic_elec.html โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/vic_electricity.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ vic_electricity_classification โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Hourly electricity demand for Victoria, Australia classified into three categories: 'low', 'medium' and 'high' โ โ according to the 20th and 80th percentiles. The dataset also includes temperature, holiday indicator and hour โ โ of the day as features. โ โ โ โ Source: โ โ O'Hara-Wild M, Hyndman R, Wang E, Godahewa R (2022). tsibbledata: Diverse Datasets for 'tsibble'. โ โ https://tsibbledata.tidyverts.org/, https://github.com/tidyverts/tsibbledata/. โ โ https://tsibbledata.tidyverts.org/reference/vic_elec.html โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/vic_electricity_classification.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ website_visits โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Daily visits to the cienciadedatos.net website registered with the google analytics service. โ โ โ โ Source: โ โ Amat Rodrigo, J. (2021). cienciadedatos.net (1.0.0). Zenodo. https://doi.org/10.5281/zenodo.10006330 โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/visitas_por_dia_web_cienciadedatos.c โ โ sv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ wikipedia_visits โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ โ Description: โ โ Log daily page views for the Wikipedia page for Peyton Manning. Scraped data using the Wikipediatrend package โ โ in R. โ โ โ โ Source: โ โ https://github.com/facebook/prophet/blob/main/examples/example_wp_log_peyton_manning.csv โ โ โ โ URL: โ โ https://raw.githubusercontent.com/skforecast/skforecast-datasets/main/data/wikipedia_visits.csv โ โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ