tasks/datasets_catalog.csv

Benchmark_task,UID,Domain_Frequency,task_name,Updated Variable Name,Variable Unit,License,Dataset Name (Metadata),Description (Metadata),Citation (Metadata),Source Link (Metadata),Preprocessing
Univariate(Economics_Weekly),federal_funds_weeks_univariate,Economics_Weekly,univariate_economics_weekly_federal_funds_effective_rate,['federal_funds_effective_rate_us'],['Percent'],Public Domain,Federal Funds Effective Rate,"The federal funds rate is the interest rate at which depository institutions trade federal funds (balances held at Federal Reserve Banks) with each other overnight. When a depository institution has surplus balances in its reserve account, it lends to other banks in need of larger balances. In simpler terms, a bank with excess cash, which is often referred to as liquidity, will lend to another bank that needs to quickly raise liquidity. The weighted average rate for these overnight negotiations is the effective federal funds rate. The Federal Open Market Committee (FOMC) sets a target for the federal funds rate and influences it through open market operations.","Board of Governors of the Federal Reserve System (US), Federal Funds Effective Rate [FF], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/FF.",https://fred.stlouisfed.org/series/FF,"Load raw file FF.csv from raw_datasets/univariate/univariate_economics_weekly_federal_funds_effective_rate/. Rename observation_date to timestamps; parse as UTC datetime. Keep single value column FF; no row filtering. Frequency check: pd.infer_freq yields W-WED (weekly, Wednesday anchor). Fill missing: drop duplicate timestamps, reindex to full W-WED grid from min to max; missing steps become NaN. Transform: one output row with variable_name=federal_funds_effective_rate_us, variable_unit=Percent. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate). Write univariate_economics_weekly_federal_funds_effective_rate.csv."
Univariate(Energy_Monthly),electricity_energy_univariate,Energy_Monthly,univariate_energy_monthly_electricity_price_kwh,['electricity_price_kwh_us_city_avg'],['U.S. Dollars'],Public Domain,Average Price: Electricity per Kilowatt-Hour in U.S. City Average,"All electricity. Average consumer prices are calculated for household fuel, motor fuel, and food items from prices collected for the Consumer Price Index (CPI). Average prices are best used to measure the price level in a particular month, not to measure price change over time. It is more appropriate to use CPI index values for the particular item categories to measure price change. Prices, except for electricity, are collected monthly by BLS representatives in the 75 urban areas priced for the CPI. Electricity prices are collected for the BLS for the same 75 areas on a monthly basis by the Department of Energy using mail questionnaires. All fuel prices include applicable Federal, State, and local taxes; prices for natural gas and electricity also include fuel and purchased gas adjustments.","U.S. Bureau of Labor Statistics, Average Price: Electricity per Kilowatt-Hour in U.S. City Average [APU000072610], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/APU000072610, August 29, 2025.",https://fred.stlouisfed.org/series/APU000072610,"Load APU000072610.csv; rename observation_date to timestamps (UTC). Keep value column APU000072610 only. Frequency check: MS (month start). Fill missing: standard monthly reindex on MS grid. Transform: variable_name=electricity_price_kwh_us_city_avg, variable_unit=U.S. Dollars. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Climate_Daily),delhi_climate_univariate,Climate_Daily,univariate_climate_daily_mean_humidity_delhi,['mean_humidity_delhi_india'],['g/m³'],Other (API Terms),Daily Humidity Time Series Data - Delhi,"The Dataset is fully dedicated for the developers who want to train the model on Weather Forecasting for Indian climate. This dataset provides data from 1st January 2013 to 24th April 2017 in the city of Delhi, India. The 1 parameter here is humidity.","SumanthVrao. “Daily Climate Time Series Data.” Kaggle, uploaded by SumanthVrao, https://www.kaggle.com/datasets/sumanthvrao/daily-climate-time-series-data.",https://www.kaggle.com/datasets/sumanthvrao/daily-climate-time-series-data,"Load DailyDelhiClimateTrain.csv; rename date to timestamps (UTC). Raw columns include meantemp, humidity, wind_speed, meanpressure; export only humidity. Removed from output variates: meantemp, wind_speed, meanpressure. Frequency check: D (daily). Fill missing: standard daily reindex. Transform: variable_name=mean_humidity_delhi_india, variable_unit=g/m³. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Software_Daily),sw_job_postings_software_univariate,Software_Daily,univariate_software_daily_software_dev_postings_indeed,['software_dev_postings_us_indeed'],"['Index Feb 1, 2020=100, Count']",Proprietary (Indeed),Software Development Job Postings on Indeed in the United States,"Indeed calculates the index change in seasonally-adjusted job postings since February 1, 2020, the pre-pandemic baseline. Indeed seasonally adjusts each series based on historical patterns in 2017, 2018, and 2019. Each series, including the national trend, occupational sectors, and sub-national geographies, is seasonally adjusted separately. Indeed switched to this new methodology in December 2022 and now reports all historical data using this new methodology. Historical numbers have been revised and may differ significantly from originally reported values. The new methodology applies a detrended seasonal adjustment factor to the index change in job postings. For more information, see Frequently Asked Questions regarding Indeed Data.","Indeed, Software Development Job Postings on Indeed in the United States [IHLIDXUSTPSOFTDEVE], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/IHLIDXUSTPSOFTDEVE, April 28, 2026.",https://fred.stlouisfed.org/series/IHLIDXUSTPSOFTDEVE,"Load IHLIDXUSTPSOFTDEVE (1).csv; rename observation_date to timestamps (UTC). Keep value column IHLIDXUSTPSOFTDEVE only. Frequency check: D (daily). Fill missing: standard daily reindex. Transform: variable_name=software_dev_postings_us_indeed, variable_unit=Index Feb 1, 2020=100, Count. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Web_Hourly),web_traffic_univariate,Web_Hourly,univariate_web_hourly_hourly_web_requests,['hourly_web_requests_single_website'],['Count'],CC0: Public Domain,Web Traffic Dataset - Single Website,"A time series containing observations of the total number of web requests (traffic counts) to a single website. The data is recorded at an hourly (and occasionally half-hourly) frequency, beginning on January 20, 2020.","RaminHuseyn. “Web Traffic Time Series Dataset.” Kaggle, uploaded by RaminHuseyn, updated a year ago, https://www.kaggle.com/datasets/raminhuseyn/web-traffic-time-series-dataset.",https://www.kaggle.com/datasets/raminhuseyn/web-traffic-time-series-dataset,"Load web_traffic.csv; rename Timestamp to timestamps (UTC). Keep value column TrafficCount only. Frequency check: median interval ~3600s suggests hourly, but fill uses a custom grid. Fill missing (special): drop all rows at exactly 00:00:00 (midnight boundary artifacts); reindex to strict 30min grid (:00 and :30 each hour); missing half-hours become NaN. Transform: variable_name=hourly_web_requests_single_website, variable_unit=Count. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Sales_Quarterly),german_houses_sales_univariate,Sales_Quarterly,univariate_sales_quarterly_real_residential_property_prices,['real_residential_property_prices_quaterly_germany'],"['Index 2010=100, U.S. Dollar']",Open Data (BIS),Real Residential Property Prices for Germany,"Quarterly observations of the real (CPI-deflated) residential property price index (2010=100) for Germany. It is a ""selected"" BIS indicator covering all types of owner-occupied new and existing dwellings across the whole country. The data is not seasonally adjusted and contains historical series backdated to 1970.","Bank for International Settlements, Real Residential Property Prices for Germany [QDER628BIS], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/QDER628BIS, April 28, 2026.",https://fred.stlouisfed.org/series/QDER628BIS,"Load QDER628BIS.csv; rename observation_date to timestamps (UTC). Keep value column QDER628BIS only. Frequency check: AS-JAN (annual, January year-start anchor) — one observation per year despite quarterly folder naming. Fill missing: standard annual reindex on AS-JAN grid. Transform: variable_name=real_residential_property_prices_quaterly_germany, variable_unit=Index 2010=100, U.S. Dollar. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Nature_Minutes),soil_nature_univariate,Nature_Minutes,univariate_nature_minutes_soil_moisture,['soil_moisture_env_monitoring_2_min'],['Percent (%)'],CC0: Public Domain,Soil and Environmental Monitoring - Soil Moisture,"This dataset provides a detailed time-series record of soil moisture data collected over a specific period. The data is structured to facilitate analysis of environmental conditions and their impact on soil properties. This dataset is particularly useful for researchers, environmental scientists, and agricultural professionals interested in understanding soil health, vegetation monitoring, and environmental modeling.","NoeyIsLearning. “Soil and Environmental Monitoring.” Kaggle, uploaded by NoeyIsLearning, updated 8 months ago, https://www.kaggle.com/datasets/noeyislearning/soil-and-environmental-monitoring.",https://www.kaggle.com/datasets/noeyislearning/soil-and-environmental-monitoring,"Load soil_moisture.csv (~129 columns: spectral bands + environmental sensors); rename datetime to timestamps (UTC). Export only soil_moisture; all other raw columns omitted from output. Frequency check: median interval ~120s; infer_freq fallback would be min. Fill missing (special): force strict 2min grid reindex (not 1min). Transform: variable_name=soil_moisture_env_monitoring_2_min, variable_unit=Percent (%). Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Healthcare_Monthly),employees_healthcare_univariate,Healthcare_Monthly,univariate_healthcare_monthly_all_employees_health_care,['all_employees_health_care_us_bls'],['Thousands of Persons'],Public Domain,All Employees in Health Care - U.S. Bureau of Labor Statistics,"Monthly observations tracking the total number of employees in the health care sector in the United States, measured in thousands of persons. The data comes from the Current Employment Statistics (Establishment Survey) and is seasonally adjusted.","U.S. Bureau of Labor Statistics, All Employees, Health Care [CES6562000101], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/CES6562000101, August 28, 2025.",https://fred.stlouisfed.org/series/CES6562000101,"Load CES6562000101.csv; rename observation_date to timestamps (UTC). Keep value column CES6562000101 only. Frequency check: MS (month start). Fill missing: standard monthly reindex. Transform: variable_name=all_employees_health_care_us_bls, variable_unit=Thousands of Persons. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Manufacturing_Seconds),manufacturing_seconds_univariate,Manufacturing_Seconds,univariate_manufacturing_seconds_skab_pressure,['Skoltech Anomaly Benchmark (SKAB) - anomaly-free - Pressure'],['Bar'],Apache 2.0 / CC BY 4.0,Skoltech Anomaly Benchmark (SKAB),"The Skoltech Anomaly Benchmark (SKAB) is designed for evaluating algorithms in industrial IoT using sensor data from a physical water-circulation testbed. Measurements are recorded at 1-second intervals. For this univariate dataset, we are using only the pressure column from the anomaly-free.csv file.","Katser, Y., Kozitsin, V., Maksimov, V., & Martynenko, I. (2021). Skoltech Anomaly Benchmark (SKAB) [Data set]. Kaggle. https://www.kaggle.com/datasets/yuriykatser/skoltech-anomaly-benchmark-skab",https://www.kaggle.com/datasets/yuriykatser/skoltech-anomaly-benchmark-skab,"Load anomaly-free (1).csv with sep="";""; rename datetime to timestamps (UTC). Raw file has 8 sensor columns; export only Pressure. Removed from output: Accelerometer1RMS, Accelerometer2RMS, Current, Temperature, Thermocouple, Voltage, Volume Flow RateRMS. Frequency check: s (1-second; median interval 1s). Fill missing (special): force strict 1-second grid reindex. Transform: variable_name=Skoltech Anomaly Benchmark (SKAB) - anomaly-free - Pressure, variable_unit=Bar. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Univariate(Transport_Annually),us_air_passengers_transport_univariate,Transport_Annually,univariate_transport_annually_air_passengers_carried,"['Air transport, passengers carried - United States']",['Count'],CC BY 4.0,"Air transport, passengers carried","Annual observations tracking the total number of domestic and international aircraft passengers of air carriers registered in the country. For this univariate dataset, we are extracting and using only the historical data specific to the United States.","World Bank Group. Air transport, passengers carried [Data set]. International Civil Aviation Organization, Civil Aviation Statistics of the World and ICAO staff estimates. https://data.worldbank.org/indicator/IS.AIR.PSGR",https://data.worldbank.org/indicator/IS.AIR.PSGR?end=2023&start=1970&view=chart,"Load API_IS.AIR.PSGR_DS2_en_csv_v2_115540.csv with skiprows=4 (World Bank wide format). Filter to Country Name == United States; melt year columns (1960–2023) to long format. Construct timestamps as {year}-12-31 UTC; value column passengers; drop rows with NaN passengers. Removed: all other countries and wide year columns after melt. Frequency check: A-DEC (annual, December anchor). Fill missing: standard annual reindex. Transform: variable_name=Air transport, passengers carried - United States, variable_unit=Count. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Energy_Minutes),split_smart_energy_multivariate,Energy_Minutes,multivariate_energy_minutes_splitsmart_ac_energy,"['Current', 'voltage', 'Power_Factor', 'Real_Power', 'room_temp', 'External_temp', 'Humidity', 'unit_consumption']","['Amperes', 'Volts', 'Dimensionless', 'Watts', 'Celsius', 'Celsius', 'Percentage', 'kWh']",CC BY 4.0,SplitSmart: An Open Dataset for Enabling Research in Energy-Efficient Ductless-Split Air Conditioner,"IoT sensor data collected over four years from ductless-split AC units in a living lab setting at BITS Pilani. It combines electrical consumption metrics (power factor, voltage, current) with environmental factors (internal/external temperature, humidity) to facilitate machine learning research in optimizing HVAC energy efficiency and reducing carbon footprints.","Kaushik, K. (2023). SplitSmart: An Open Dataset for Enabling Research in Energy-Efficient Ductless-Split Air Conditioner [Data set]. Data.gov.",https://catalog.data.gov/dataset/splitsmart-an-open-dataset-for-enabling-research-in-energy-efficient-ductless-split-air-co,"Load dataA200AC03.csv (device A200AC03, room A200 third AC unit). Kept columns: Time_Stamp + Current, voltage, Power_Factor, Real_Power, room_temp, External_temp, Humidity, unit_consumption. Removed metadata/group columns: _id, Device_ID, _updated, _created, _etag. Rename Time_Stamp to timestamps (UTC). Frequency check (special): median interval ~121s (~2 min, not 1 min); floor timestamps to 2min; freq=2min. Fill missing (special): drop duplicate timestamps; reindex to strict 2min grid; missing steps NaN. Transform: 8 variates (Current, voltage, Power_Factor, Real_Power, room_temp, External_temp, Humidity, unit_consumption); variable_unit blank per variate. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Energy_Minutes),split_smart_energy_multivariate_covariate,Energy_Minutes,covariate_energy_minutes_splitsmart_ac_energy,['unit_consumption'],['kWh'],CC BY 4.0,SplitSmart: An Open Dataset for Enabling Research in Energy-Efficient Ductless-Split Air Conditioner,"IoT sensor data collected over four years from ductless-split AC units in a living lab setting at BITS Pilani. It combines electrical consumption metrics (power factor, voltage, current) with environmental factors (internal/external temperature, humidity) to facilitate machine learning research in optimizing HVAC energy efficiency and reducing carbon footprints.","Kaushik, K. (2023). SplitSmart: An Open Dataset for Enabling Research in Energy-Efficient Ductless-Split Air Conditioner [Data set]. Data.gov.",https://catalog.data.gov/dataset/splitsmart-an-open-dataset-for-enabling-research-in-energy-efficient-ductless-split-air-co,"Load dataA200AC03.csv (device A200AC03, room A200 third AC unit). Kept columns: Time_Stamp + Current, voltage, Power_Factor, Real_Power, room_temp, External_temp, Humidity, unit_consumption. Removed metadata/group columns: _id, Device_ID, _updated, _created, _etag. Rename Time_Stamp to timestamps (UTC). Frequency check (special): median interval ~121s (~2 min, not 1 min); floor timestamps to 2min; freq=2min. Fill missing (special): drop duplicate timestamps; reindex to strict 2min grid; missing steps NaN. Transform: 8 variates (Current, voltage, Power_Factor, Real_Power, room_temp, External_temp, Humidity, unit_consumption); variable_unit blank per variate. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Healthcare_Daily),nyc_covid_healthcare_multivariate,Healthcare_Daily,multivariate_healthcare_daily_nyc_covid_counts,"['CASE_COUNT', 'PROBABLE_CASE_COUNT', 'HOSPITALIZED_COUNT', 'DEATH_COUNT', 'CASE_COUNT_7DAY_AVG', 'ALL_CASE_COUNT_7DAY_AVG', 'HOSP_COUNT_7DAY_AVG', 'DEATH_COUNT_7DAY_AVG', 'BX_CASE_COUNT', 'BX_PROBABLE_CASE_COUNT', 'BX_HOSPITALIZED_COUNT', 'BX_DEATH_COUNT', 'BX_CASE_COUNT_7DAY_AVG', 'BX_PROBABLE_CASE_COUNT_7DAY_AVG', 'BX_ALL_CASE_COUNT_7DAY_AVG', 'BX_HOSPITALIZED_COUNT_7DAY_AVG', 'BX_DEATH_COUNT_7DAY_AVG', 'BK_CASE_COUNT', 'BK_PROBABLE_CASE_COUNT', 'BK_HOSPITALIZED_COUNT', 'BK_DEATH_COUNT', 'BK_CASE_COUNT_7DAY_AVG', 'BK_PROBABLE_CASE_COUNT_7DAY_AVG', 'BK_ALL_CASE_COUNT_7DAY_AVG', 'BK_HOSPITALIZED_COUNT_7DAY_AVG', 'BK_DEATH_COUNT_7DAY_AVG', 'MN_CASE_COUNT', 'MN_PROBABLE_CASE_COUNT', 'MN_HOSPITALIZED_COUNT', 'MN_DEATH_COUNT', 'MN_CASE_COUNT_7DAY_AVG', 'MN_PROBABLE_CASE_COUNT_7DAY_AVG', 'MN_ALL_CASE_COUNT_7DAY_AVG', 'MN_HOSPITALIZED_COUNT_7DAY_AVG', 'MN_DEATH_COUNT_7DAY_AVG', 'QN_CASE_COUNT', 'QN_PROBABLE_CASE_COUNT', 'QN_HOSPITALIZED_COUNT', 'QN_DEATH_COUNT', 'QN_CASE_COUNT_7DAY_AVG', 'QN_PROBABLE_CASE_COUNT_7DAY_AVG', 'QN_ALL_CASE_COUNT_7DAY_AVG', 'QN_HOSPITALIZED_COUNT_7DAY_AVG', 'QN_DEATH_COUNT_7DAY_AVG', 'SI_CASE_COUNT', 'SI_PROBABLE_CASE_COUNT', 'SI_HOSPITALIZED_COUNT', 'SI_DEATH_COUNT', 'SI_PROBABLE_CASE_COUNT_7DAY_AVG', 'SI_CASE_COUNT_7DAY_AVG', 'SI_ALL_CASE_COUNT_7DAY_AVG', 'SI_HOSPITALIZED_COUNT_7DAY_AVG', 'SI_DEATH_COUNT_7DAY_AVG']","['Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count', 'Count']",Public Domain (NYC Open Data),"COVID-19 Daily Counts of Cases, Hospitalizations, and Deaths","Granular epidemiological dataset tracking COVID-19 confirmed and probable cases, hospitalizations, and deaths across New York City. The data provides both city-wide aggregates and specific regional breakdowns across all five boroughs (Bronx, Brooklyn, Manhattan, Queens, Staten Island), making it an ideal benchmark for spatial-temporal infectious disease modeling.","Department of Health and Mental Hygiene (DOHMH). COVID-19 Daily Counts of Cases, Hospitalizations, and Deaths [Data set]. City of New York / Data.gov.",https://catalog.data.gov/dataset/covid-19-daily-counts-of-cases-hospitalizations-and-deaths,"Load COVID-19_Daily_Counts_of_Cases__Hospitalizations__and_Deaths.csv. Keep date_of_interest + 53 count series: CASE_COUNT, PROBABLE_CASE_COUNT, HOSPITALIZED_COUNT, DEATH_COUNT, CASE_COUNT_7DAY_AVG, ALL_CASE_COUNT_7DAY_AVG, HOSP_COUNT_7DAY_AVG, DEATH_COUNT_7DAY_AVG, BX_CASE_COUNT, BX_PROBABLE_CASE_COUNT, BX_HOSPITALIZED_COUNT, BX_DEATH_COUNT, BX_CASE_COUNT_7DAY_AVG, BX_PROBABLE_CASE_COUNT_7DAY_AVG, BX_ALL_CASE_COUNT_7DAY_AVG, BX_HOSPITALIZED_COUNT_7DAY_AVG, BX_DEATH_COUNT_7DAY_AVG, BK_CASE_COUNT, BK_PROBABLE_CASE_COUNT, BK_HOSPITALIZED_COUNT, BK_DEATH_COUNT, BK_CASE_COUNT_7DAY_AVG, BK_PROBABLE_CASE_COUNT_7DAY_AVG, BK_ALL_CASE_COUNT_7DAY_AVG, BK_HOSPITALIZED_COUNT_7DAY_AVG, BK_DEATH_COUNT_7DAY_AVG, MN_CASE_COUNT, MN_PROBABLE_CASE_COUNT, MN_HOSPITALIZED_COUNT, MN_DEATH_COUNT, MN_CASE_COUNT_7DAY_AVG, MN_PROBABLE_CASE_COUNT_7DAY_AVG, MN_ALL_CASE_COUNT_7DAY_AVG, MN_HOSPITALIZED_COUNT_7DAY_AVG, MN_DEATH_COUNT_7DAY_AVG, QN_CASE_COUNT, QN_PROBABLE_CASE_COUNT, QN_HOSPITALIZED_COUNT, QN_DEATH_COUNT, QN_CASE_COUNT_7DAY_AVG, QN_PROBABLE_CASE_COUNT_7DAY_AVG, QN_ALL_CASE_COUNT_7DAY_AVG, QN_HOSPITALIZED_COUNT_7DAY_AVG, QN_DEATH_COUNT_7DAY_AVG, SI_CASE_COUNT, SI_PROBABLE_CASE_COUNT, SI_HOSPITALIZED_COUNT, SI_DEATH_COUNT, SI_PROBABLE_CASE_COUNT_7DAY_AVG, SI_CASE_COUNT_7DAY_AVG, SI_ALL_CASE_COUNT_7DAY_AVG, SI_HOSPITALIZED_COUNT_7DAY_AVG, SI_DEATH_COUNT_7DAY_AVG. Removed column INCOMPLETE (reporting completeness flag, not a count series). Rename date_of_interest to timestamps; parse with format %m/%d/%Y (UTC). Frequency check: D (daily). Fill missing: standard daily reindex on all 53 variates. Transform: 53 rows (one per variate); variable_unit blank; _to_json_number helper for values. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Healthcare_Daily),nyc_covid_healthcare_multivariate_covariate,Healthcare_Daily,covariate_healthcare_daily_nyc_covid_counts,['DEATH_COUNT'],['Count'],Public Domain (NYC Open Data),"COVID-19 Daily Counts of Cases, Hospitalizations, and Deaths","Granular epidemiological dataset tracking COVID-19 confirmed and probable cases, hospitalizations, and deaths across New York City. The data provides both city-wide aggregates and specific regional breakdowns across all five boroughs (Bronx, Brooklyn, Manhattan, Queens, Staten Island), making it an ideal benchmark for spatial-temporal infectious disease modeling.","Department of Health and Mental Hygiene (DOHMH). COVID-19 Daily Counts of Cases, Hospitalizations, and Deaths [Data set]. City of New York / Data.gov.",https://catalog.data.gov/dataset/covid-19-daily-counts-of-cases-hospitalizations-and-deaths,"Load COVID-19_Daily_Counts_of_Cases__Hospitalizations__and_Deaths.csv. Keep date_of_interest + 53 count series: CASE_COUNT, PROBABLE_CASE_COUNT, HOSPITALIZED_COUNT, DEATH_COUNT, CASE_COUNT_7DAY_AVG, ALL_CASE_COUNT_7DAY_AVG, HOSP_COUNT_7DAY_AVG, DEATH_COUNT_7DAY_AVG, BX_CASE_COUNT, BX_PROBABLE_CASE_COUNT, BX_HOSPITALIZED_COUNT, BX_DEATH_COUNT, BX_CASE_COUNT_7DAY_AVG, BX_PROBABLE_CASE_COUNT_7DAY_AVG, BX_ALL_CASE_COUNT_7DAY_AVG, BX_HOSPITALIZED_COUNT_7DAY_AVG, BX_DEATH_COUNT_7DAY_AVG, BK_CASE_COUNT, BK_PROBABLE_CASE_COUNT, BK_HOSPITALIZED_COUNT, BK_DEATH_COUNT, BK_CASE_COUNT_7DAY_AVG, BK_PROBABLE_CASE_COUNT_7DAY_AVG, BK_ALL_CASE_COUNT_7DAY_AVG, BK_HOSPITALIZED_COUNT_7DAY_AVG, BK_DEATH_COUNT_7DAY_AVG, MN_CASE_COUNT, MN_PROBABLE_CASE_COUNT, MN_HOSPITALIZED_COUNT, MN_DEATH_COUNT, MN_CASE_COUNT_7DAY_AVG, MN_PROBABLE_CASE_COUNT_7DAY_AVG, MN_ALL_CASE_COUNT_7DAY_AVG, MN_HOSPITALIZED_COUNT_7DAY_AVG, MN_DEATH_COUNT_7DAY_AVG, QN_CASE_COUNT, QN_PROBABLE_CASE_COUNT, QN_HOSPITALIZED_COUNT, QN_DEATH_COUNT, QN_CASE_COUNT_7DAY_AVG, QN_PROBABLE_CASE_COUNT_7DAY_AVG, QN_ALL_CASE_COUNT_7DAY_AVG, QN_HOSPITALIZED_COUNT_7DAY_AVG, QN_DEATH_COUNT_7DAY_AVG, SI_CASE_COUNT, SI_PROBABLE_CASE_COUNT, SI_HOSPITALIZED_COUNT, SI_DEATH_COUNT, SI_PROBABLE_CASE_COUNT_7DAY_AVG, SI_CASE_COUNT_7DAY_AVG, SI_ALL_CASE_COUNT_7DAY_AVG, SI_HOSPITALIZED_COUNT_7DAY_AVG, SI_DEATH_COUNT_7DAY_AVG. Removed column INCOMPLETE (reporting completeness flag, not a count series). Rename date_of_interest to timestamps; parse with format %m/%d/%Y (UTC). Frequency check: D (daily). Fill missing: standard daily reindex on all 53 variates. Transform: 53 rows (one per variate); variable_unit blank; _to_json_number helper for values. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Climate_Hourly),madrid_hours_multivariate,Climate_Hourly,multivariate_climate_hourly_madrid_air_pollution,"['BEN', 'CH4', 'CO', 'EBE', 'NMHC', 'NO', 'NO2', 'NOx', 'O3', 'PM10', 'PM25', 'SO2', 'TCH', 'TOL']","['µg/m³', 'mg/m³', 'mg/m³', 'µg/m³', 'mg/m³', 'µg/m³', 'µg/m³', 'µg/m³', 'µg/m³', 'µg/m³', 'µg/m³', 'µg/m³', 'mg/m³', 'µg/m³']",CC0: Public Domain,2001-2022 Hourly dataset of pollution in Madrid,"A massive, two-decade-long atmospheric dataset collected from 24 environmental monitoring stations across Madrid. It tracks hourly concentrations of highly specific pollutants including particulate matter (PM10, PM2.5), nitrogen oxides, ozone, and sulfur dioxide. Highly valuable for identifying seasonal climate trends, daily urban traffic cycles, and the long-term impact of emission policies.",Ignacio Q.G. (2022). 2001-2022 Hourly dataset of pollution in Madrid [Data set]. Kaggle.,https://www.kaggle.com/datasets/ignacioqg/20012022-hourly-dataset-of-pollution-in-madrid,"Load MadridPolution2001-2022.csv; rename Time to timestamps (UTC). Kept 14 pollutant variates: BEN, CH4, CO, EBE, NMHC, NO, NO2, NOx, O3, PM10, PM25, SO2, TCH, TOL. All other raw columns (station metadata, etc.) excluded from VALUE_COLS. Frequency check: h (hourly). Fill missing: standard hourly reindex; sparse hourly coverage yields many NaN fills. Transform: 14 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Climate_Hourly),madrid_hours_multivariate_covariate,Climate_Hourly,covariate_climate_hourly_madrid_air_pollution,['PM25'],['µg/m³'],CC0: Public Domain,2001-2022 Hourly dataset of pollution in Madrid,"A massive, two-decade-long atmospheric dataset collected from 24 environmental monitoring stations across Madrid. It tracks hourly concentrations of highly specific pollutants including particulate matter (PM10, PM2.5), nitrogen oxides, ozone, and sulfur dioxide. Highly valuable for identifying seasonal climate trends, daily urban traffic cycles, and the long-term impact of emission policies.",Ignacio Q.G. (2022). 2001-2022 Hourly dataset of pollution in Madrid [Data set]. Kaggle.,https://www.kaggle.com/datasets/ignacioqg/20012022-hourly-dataset-of-pollution-in-madrid,"Load MadridPolution2001-2022.csv; rename Time to timestamps (UTC). Kept 14 pollutant variates: BEN, CH4, CO, EBE, NMHC, NO, NO2, NOx, O3, PM10, PM25, SO2, TCH, TOL. All other raw columns (station metadata, etc.) excluded from VALUE_COLS. Frequency check: h (hourly). Fill missing: standard hourly reindex; sparse hourly coverage yields many NaN fills. Transform: 14 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Transport_Monthly),baggage_months_multivariate,Transport_Monthly,multivariate_transport_monthly_airline_baggage_complaints,"['Baggage', 'Scheduled', 'Cancelled', 'Enplaned']","['Complaints', 'Flights', 'Flights', 'Passengers']",Public Domain,Airline Baggage Complaints,"A multivariate transport logistics dataset tracking monthly observations from 2004 to 2010 for major US airlines. By juxtaposing total passenger complaints regarding lost or damaged baggage against total scheduled flights, cancellations, and boarded passengers, this data allows models to forecast operational friction and its subsequent impact on customer satisfaction in the aviation industry.","Santello, G. (2021). Airline Baggage Complaints - Time Series Dataset [Data set]. Kaggle.",https://www.kaggle.com/datasets/gabrielsantello/airline-baggage-complaints-time-series-dataset,"Load baggagecomplaints.csv. Filter to first airline in file (American Eagle) via df['Airline'].iloc[0]. Parse Date with format %m/%Y as UTC month-start timestamps. Kept variates: Baggage, Scheduled, Cancelled, Enplaned. Non-exported columns during fill frame: Airline, Month, Year (metadata; not in transform output). Frequency check: MS (month start). Fill missing: standard monthly reindex. Transform: 4 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Transport_Monthly),baggage_months_multivariate_covariate,Transport_Monthly,covariate_transport_monthly_airline_baggage_complaints,['Baggage'],['Complaints'],Public Domain,Airline Baggage Complaints,"A multivariate transport logistics dataset tracking monthly observations from 2004 to 2010 for major US airlines. By juxtaposing total passenger complaints regarding lost or damaged baggage against total scheduled flights, cancellations, and boarded passengers, this data allows models to forecast operational friction and its subsequent impact on customer satisfaction in the aviation industry.","Santello, G. (2021). Airline Baggage Complaints - Time Series Dataset [Data set]. Kaggle.",https://www.kaggle.com/datasets/gabrielsantello/airline-baggage-complaints-time-series-dataset,"Load baggagecomplaints.csv. Filter to first airline in file (American Eagle) via df['Airline'].iloc[0]. Parse Date with format %m/%Y as UTC month-start timestamps. Kept variates: Baggage, Scheduled, Cancelled, Enplaned. Non-exported columns during fill frame: Airline, Month, Year (metadata; not in transform output). Frequency check: MS (month start). Fill missing: standard monthly reindex. Transform: 4 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Sales_Weekly),advertising_sales_covariate,Sales_Weekly,multivariate_sales_weekly_walmart_sales,"['Weekly_Sales', 'Holiday_Flag', 'Temperature', 'Fuel_Price', 'CPI', 'Unemployment']","['USD', 'Binary', 'Fahrenheit', 'USD/Gallon', 'Index', 'Percentage']",Other (Proprietary),Walmart Sales,"Comprehensive retail dataset capturing weekly revenue figures across multiple stores, heavily enriched with external covariates. Models must leverage regional environmental factors (air temperature), logistical costs (fuel prices), macroeconomic indicators (unemployment, CPI), and seasonal spikes (holiday flags) to accurately predict purchasing power and consumer demand.",Mikhail. (2022). Walmart Sales [Data set]. Kaggle.,https://www.kaggle.com/datasets/mikhail1681/walmart-sales,"Load Walmart_Sales.csv; filter to Store == 1. Parse Date with format %d-%m-%Y (day-first) as UTC. Kept variates: Weekly_Sales, Holiday_Flag, Temperature, Fuel_Price, CPI, Unemployment. Store column retained in fill frame but not exported as a variate. Frequency check: W-FRI (weekly, Friday anchor). Fill missing: standard weekly reindex. Transform: 6 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Sales_Weekly),advertising_sales_covariate,Sales_Weekly,covariate_sales_weekly_walmart_sales,['Weekly_Sales'],['USD'],Other (Proprietary),Walmart Sales,"Comprehensive retail dataset capturing weekly revenue figures across multiple stores, heavily enriched with external covariates. Models must leverage regional environmental factors (air temperature), logistical costs (fuel prices), macroeconomic indicators (unemployment, CPI), and seasonal spikes (holiday flags) to accurately predict purchasing power and consumer demand.",Mikhail. (2022). Walmart Sales [Data set]. Kaggle.,https://www.kaggle.com/datasets/mikhail1681/walmart-sales,"Load Walmart_Sales.csv; filter to Store == 1. Parse Date with format %d-%m-%Y (day-first) as UTC. Kept variates: Weekly_Sales, Holiday_Flag, Temperature, Fuel_Price, CPI, Unemployment. Store column retained in fill frame but not exported as a variate. Frequency check: W-FRI (weekly, Friday anchor). Fill missing: standard weekly reindex. Transform: 6 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Software_Hourly),cybersecurity_software_covariate,Software_Hourly,multivariate_software_hourly_cybertec_iiot_malware,"['Packet Size', 'Packet Length', 'Inter-Arrival Time', 'Flow Duration', 'Total Packets', 'Total Bytes', 'Average Packet Size', 'Packet Arrival Rate', 'Payload Entropy', 'Flow Entropy', 'Baseline Deviation', 'Packet Size Variance', 'Malicious Signatures', 'Embedded Commands', 'Device Activity Patterns', 'CPU Usage', 'Memory Usage', 'Network Interface Stats', 'Duration Anomaly', 'Known IoC', 'C&C Communication', 'Data Exfiltration']","['Bytes', 'Bytes', 'Seconds', 'Seconds', 'Count', 'Bytes', 'Bytes', 'Packets/sec', 'Shannon Entropy', 'Shannon Entropy', 'Percentage', 'Bytes²', 'Count', 'Count', 'Categorical', 'Percentage', 'Percentage', 'Bytes/sec', 'Seconds', 'Binary', 'Binary', 'Binary']",CC BY-NC-SA 4.0,CyberTec IIoT Malware Dataset,"A cutting-edge cybersecurity dataset capturing hourly network traffic and system health logs across an Industrial Internet of Things (IIoT) environment. It merges low-level network behaviors (packet entropy, flow duration) with system resource utilization (CPU/Memory) to facilitate AI-driven classification of zero-day intrusions, ransomware, botnets, and data exfiltration attempts.",DatasetEngineer. (2024). CyberTec IIoT Malware Dataset (CIMD-2024) [Data set]. Kaggle.,https://www.kaggle.com/datasets/datasetengineer/cybertec-iiot-malware-dataset-cimd-2024,"Load IIoT_Malware_Timeseries_Dataset.csv; rename Timestamp to timestamps (UTC). Excluded columns (15): Timestamp, Source IP, Destination IP, Source Port, Destination Port, Time of Day, Day of Week, Protocol Type, Flags, Payload Pattern, Device Type, Label, Attack Type, Device Context, Threat Intensity. Excluded categorical label/leakage columns: Label, Attack Type, Device Context, Threat Intensity, Protocol Type, Flags, Payload Pattern, Device Type. Excluded identifier/redundant columns: Source IP, Destination IP, Source Port, Destination Port, Time of Day, Day of Week. Kept 22 numeric variates: Packet Size, Packet Length, Inter-Arrival Time, Flow Duration, Total Packets, Total Bytes, Average Packet Size, Packet Arrival Rate, Payload Entropy, Flow Entropy, Baseline Deviation, Packet Size Variance, Malicious Signatures, Embedded Commands, Device Activity Patterns, CPU Usage, Memory Usage, Network Interface Stats, Duration Anomaly, Known IoC, C&C Communication, Data Exfiltration. Frequency check: H (hourly). Fill missing: standard hourly reindex on value columns. Transform: 22 variates; variable_unit blank; _to_json_number helper. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Software_Hourly),cybersecurity_software_covariate,Software_Hourly,covariate_software_hourly_cybertec_iiot_malware,['Malicious Signatures'],['Count'],CC BY-NC-SA 4.0,CyberTec IIoT Malware Dataset,"A cutting-edge cybersecurity dataset capturing hourly network traffic and system health logs across an Industrial Internet of Things (IIoT) environment. It merges low-level network behaviors (packet entropy, flow duration) with system resource utilization (CPU/Memory) to facilitate AI-driven classification of zero-day intrusions, ransomware, botnets, and data exfiltration attempts.",DatasetEngineer. (2024). CyberTec IIoT Malware Dataset (CIMD-2024) [Data set]. Kaggle.,https://www.kaggle.com/datasets/datasetengineer/cybertec-iiot-malware-dataset-cimd-2024,"Load IIoT_Malware_Timeseries_Dataset.csv; rename Timestamp to timestamps (UTC). Excluded columns (15): Timestamp, Source IP, Destination IP, Source Port, Destination Port, Time of Day, Day of Week, Protocol Type, Flags, Payload Pattern, Device Type, Label, Attack Type, Device Context, Threat Intensity. Excluded categorical label/leakage columns: Label, Attack Type, Device Context, Threat Intensity, Protocol Type, Flags, Payload Pattern, Device Type. Excluded identifier/redundant columns: Source IP, Destination IP, Source Port, Destination Port, Time of Day, Day of Week. Kept 22 numeric variates: Packet Size, Packet Length, Inter-Arrival Time, Flow Duration, Total Packets, Total Bytes, Average Packet Size, Packet Arrival Rate, Payload Entropy, Flow Entropy, Baseline Deviation, Packet Size Variance, Malicious Signatures, Embedded Commands, Device Activity Patterns, CPU Usage, Memory Usage, Network Interface Stats, Duration Anomaly, Known IoC, C&C Communication, Data Exfiltration. Frequency check: H (hourly). Fill missing: standard hourly reindex on value columns. Transform: 22 variates; variable_unit blank; _to_json_number helper. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Manufacturing_Seconds),skab_manufacturing_multivariate,Manufacturing_Seconds,multivariate_manufacturing_seconds_skab,"['Accelerometer1RMS', 'Accelerometer2RMS', 'Current', 'Pressure', 'Temperature', 'Thermocouple', 'Voltage', 'Volume Flow RateRMS']","['g', 'g', 'Ampere', 'Bar', 'Celsius', 'Celsius', 'Volt', 'L/min']",Apache 2.0 / CC BY 4.0,Skoltech Anomaly Benchmark (SKAB),"A high-frequency physical testbed dataset specifically engineered for predictive maintenance and industrial anomaly detection. It tracks continuous 1-second telemetry from eight distinct sensors—including vibrations, pressure dynamics, and volume flow rates—mounted on a functioning water-circulation loop, capturing both steady-state operations and mechanically injected faults.","Katser, Y., Kozitsin, V., Maksimov, V., & Martynenko, I. (2021). Skoltech Anomaly Benchmark (SKAB) [Data set]. Kaggle.",https://www.kaggle.com/datasets/yuriykatser/skoltech-anomaly-benchmark-skab,"Load anomaly-free (2).csv with sep="";""; rename datetime to timestamps (UTC). Kept all 8 sensor variates: Accelerometer1RMS, Accelerometer2RMS, Current, Pressure, Temperature, Thermocouple, Voltage, Volume Flow RateRMS. Frequency check: s (1-second). Fill missing (special): force strict 1-second grid reindex. Transform: 8 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Manufacturing_Seconds),skab_manufacturing_multivariate_covariate,Manufacturing_Seconds,covariate_manufacturing_seconds_skab,['Pressure'],['Bar'],Apache 2.0 / CC BY 4.0,Skoltech Anomaly Benchmark (SKAB),"A high-frequency physical testbed dataset specifically engineered for predictive maintenance and industrial anomaly detection. It tracks continuous 1-second telemetry from eight distinct sensors—including vibrations, pressure dynamics, and volume flow rates—mounted on a functioning water-circulation loop, capturing both steady-state operations and mechanically injected faults.","Katser, Y., Kozitsin, V., Maksimov, V., & Martynenko, I. (2021). Skoltech Anomaly Benchmark (SKAB) [Data set]. Kaggle.",https://www.kaggle.com/datasets/yuriykatser/skoltech-anomaly-benchmark-skab,"Load anomaly-free (2).csv with sep="";""; rename datetime to timestamps (UTC). Kept all 8 sensor variates: Accelerometer1RMS, Accelerometer2RMS, Current, Pressure, Temperature, Thermocouple, Voltage, Volume Flow RateRMS. Frequency check: s (1-second). Fill missing (special): force strict 1-second grid reindex. Transform: 8 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Web_Daily),wikipedia_web_multivariate,Web_Daily,multivariate_web_daily_wikipedia_web_traffic,"['2NE1_zh.wikipedia.org_all-access_spider', '2PM_zh.wikipedia.org_all-access_spider', '3C_zh.wikipedia.org_all-access_spider', '4minute_zh.wikipedia.org_all-access_spider', '52_Hz_I_Love_You_zh.wikipedia.org_all-access_spider', '5566_zh.wikipedia.org_all-access_spider', '91Days_zh.wikipedia.org_all-access_spider', ""A'N'D_zh.wikipedia.org_all-access_spider"", 'AKB48_zh.wikipedia.org_all-access_spider', 'ASCII_zh.wikipedia.org_all-access_spider', 'ASTRO_zh.wikipedia.org_all-access_spider', 'Ahq_e-Sports_Club_zh.wikipedia.org_all-access_spider', 'All_your_base_are_belong_to_us_zh.wikipedia.org_all-access_spider', 'AlphaGo_zh.wikipedia.org_all-access_spider', 'Android_zh.wikipedia.org_all-access_spider', 'Angelababy_zh.wikipedia.org_all-access_spider', 'Apink_zh.wikipedia.org_all-access_spider', 'Apple_II_zh.wikipedia.org_all-access_spider', 'As_One_zh.wikipedia.org_all-access_spider', 'B-PROJECT_zh.wikipedia.org_all-access_spider', 'B1A4_zh.wikipedia.org_all-access_spider', 'BDSM_zh.wikipedia.org_all-access_spider', 'BEAST_zh.wikipedia.org_all-access_spider', 'BIGBANG_zh.wikipedia.org_all-access_spider', 'BLACK_PINK_zh.wikipedia.org_all-access_spider', 'BLEACH_zh.wikipedia.org_all-access_spider', 'BTOB_zh.wikipedia.org_all-access_spider', 'Beautiful_Mind_zh.wikipedia.org_all-access_spider', 'Beyond_zh.wikipedia.org_all-access_spider', 'Big_zh.wikipedia.org_all-access_spider', 'Block_B_zh.wikipedia.org_all-access_spider', 'CHAOS;CHILD_zh.wikipedia.org_all-access_spider', 'CHAOS;HEAD_zh.wikipedia.org_all-access_spider', 'CLC_zh.wikipedia.org_all-access_spider', 'CNBLUE_zh.wikipedia.org_all-access_spider', 'DIA_zh.wikipedia.org_all-access_spider', 'Daigo_zh.wikipedia.org_all-access_spider', 'Dear_My_Friends_zh.wikipedia.org_all-access_spider', 'Doctors_zh.wikipedia.org_all-access_spider', 'EGOIST_zh.wikipedia.org_all-access_spider', 'EXID_zh.wikipedia.org_all-access_spider', 'EXO_zh.wikipedia.org_all-access_spider', 'Energy_zh.wikipedia.org_all-access_spider', 'Eric_Nam_zh.wikipedia.org_all-access_spider', 'FAIRY_TAIL_zh.wikipedia.org_all-access_spider', 'FIESTAR_zh.wikipedia.org_all-access_spider', 'FIRST_CLASS_zh.wikipedia.org_all-access_spider', 'Facebook_zh.wikipedia.org_all-access_spider', 'Fantastic_Duo_zh.wikipedia.org_all-access_spider', 'Fate/Grand_Order_zh.wikipedia.org_all-access_spider', 'Fate/Zero_zh.wikipedia.org_all-access_spider', 'Fate/stay_night_zh.wikipedia.org_all-access_spider', 'File:427FW_126.JPG_zh.wikipedia.org_all-access_spider', 'File:Ap_F23_20110324114153669.jpg_zh.wikipedia.org_all-access_spider', 'File:Christian-krohg-leiv-eriksson.jpg_zh.wikipedia.org_all-access_spider', 'File:Sebastiaosalgado2006.jpg_zh.wikipedia.org_all-access_spider', 'File:Zhongxiao_West_Road2.JPG_zh.wikipedia.org_all-access_spider', 'G-Dragon_zh.wikipedia.org_all-access_spider', 'G.NA_zh.wikipedia.org_all-access_spider', 'GFRIEND_zh.wikipedia.org_all-access_spider', 'GNZ48_zh.wikipedia.org_all-access_spider', 'GOT7_zh.wikipedia.org_all-access_spider', 'GUMMY_zh.wikipedia.org_all-access_spider', ""Girl's_Day_zh.wikipedia.org_all-access_spider"", 'Goodbye_Mr._Black_zh.wikipedia.org_all-access_spider', 'Google_zh.wikipedia.org_all-access_spider', 'Gu9udan_zh.wikipedia.org_all-access_spider', 'HDMI_zh.wikipedia.org_all-access_spider', 'Hand_Shakers_zh.wikipedia.org_all-access_spider', 'Hisasi_zh.wikipedia.org_all-access_spider', 'Hit_The_Stage_zh.wikipedia.org_all-access_spider', 'Hotel_King_zh.wikipedia.org_all-access_spider', ""I'm_Home_zh.wikipedia.org_all-access_spider"", 'I.O.I_zh.wikipedia.org_all-access_spider', 'IKON_zh.wikipedia.org_all-access_spider', 'INFINITE_zh.wikipedia.org_all-access_spider', 'IPhone_zh.wikipedia.org_all-access_spider', 'Ingress_zh.wikipedia.org_all-access_spider', 'Intel740_zh.wikipedia.org_all-access_spider', 'Intel_80386_zh.wikipedia.org_all-access_spider', 'Juksy_zh.wikipedia.org_all-access_spider', 'K.A.R.D_zh.wikipedia.org_all-access_spider', 'KNK_zh.wikipedia.org_all-access_spider', 'KUROMUKURO_zh.wikipedia.org_all-access_spider', 'Kara_zh.wikipedia.org_all-access_spider', 'Kill_Me_Heal_Me_zh.wikipedia.org_all-access_spider', 'Lady_Gaga_zh.wikipedia.org_all-access_spider', 'Legal_high_zh.wikipedia.org_all-access_spider', 'LoveLive!_zh.wikipedia.org_all-access_spider', 'LoveLive!_Sunshine!!_zh.wikipedia.org_all-access_spider', 'Lovelyz_zh.wikipedia.org_all-access_spider', 'Lulu_zh.wikipedia.org_all-access_spider', 'MADTOWN_zh.wikipedia.org_all-access_spider', 'MAMAMOO_zh.wikipedia.org_all-access_spider', 'MONSTA_X_zh.wikipedia.org_all-access_spider', 'Madame_Antoine_zh.wikipedia.org_all-access_spider', 'Mamamoo_zh.wikipedia.org_all-access_spider', 'Mike_D._Angelo_zh.wikipedia.org_all-access_spider', 'Miss_A_zh.wikipedia.org_all-access_spider', 'Missing9_zh.wikipedia.org_all-access_spider', 'Monstar_zh.wikipedia.org_all-access_spider', 'Mrs._Cop_2_zh.wikipedia.org_all-access_spider', 'NBA_zh.wikipedia.org_all-access_spider', 'NCT_zh.wikipedia.org_all-access_spider', 'NEW_GAME!_zh.wikipedia.org_all-access_spider', 'Netflix_zh.wikipedia.org_all-access_spider', 'Niantic_zh.wikipedia.org_all-access_spider', 'ONE_OK_ROCK_zh.wikipedia.org_all-access_spider', 'ONE_PIECE_zh.wikipedia.org_all-access_spider', 'ONE_PIECE_FILM_GOLD_zh.wikipedia.org_all-access_spider', 'Oh_My_Girl_zh.wikipedia.org_all-access_spider', 'Oh_My_Venus_zh.wikipedia.org_all-access_spider', 'PRODUCE_101_zh.wikipedia.org_all-access_spider', 'Page_Turner_zh.wikipedia.org_all-access_spider', 'Pen-Pineapple-Apple-Pen_zh.wikipedia.org_all-access_spider', 'PewDiePie_zh.wikipedia.org_all-access_spider', 'Popu_Lady_zh.wikipedia.org_all-access_spider', 'Pristin_zh.wikipedia.org_all-access_spider', 'Python_zh.wikipedia.org_all-access_spider', 'Qualidea_Code_zh.wikipedia.org_all-access_spider', 'RADWIMPS_zh.wikipedia.org_all-access_spider', 'RAID_zh.wikipedia.org_all-access_spider', 'RWBY_zh.wikipedia.org_all-access_spider', 'Rain_zh.wikipedia.org_all-access_spider', 'ReLIFE_zh.wikipedia.org_all-access_spider', 'Red_Velvet_zh.wikipedia.org_all-access_spider', 'Rewrite_zh.wikipedia.org_all-access_spider', 'Running_Man_zh.wikipedia.org_all-access_spider', 'S.H.E_zh.wikipedia.org_all-access_spider', 'SF9_zh.wikipedia.org_all-access_spider', 'SHINee_zh.wikipedia.org_all-access_spider', 'SISTAR_zh.wikipedia.org_all-access_spider', 'SIXTEEN_zh.wikipedia.org_all-access_spider', 'SMAP_zh.wikipedia.org_all-access_spider', 'SNH48_zh.wikipedia.org_all-access_spider', 'SUPER_LOVERS_zh.wikipedia.org_all-access_spider', 'Schwarzesmarken_zh.wikipedia.org_all-access_spider', 'Secret_Love_zh.wikipedia.org_all-access_spider', 'SpeXial_zh.wikipedia.org_all-access_spider', 'Special:Search_zh.wikipedia.org_all-access_spider', 'Super_Junior_zh.wikipedia.org_all-access_spider', 'Supper_Moment_zh.wikipedia.org_all-access_spider', 'T-ara_zh.wikipedia.org_all-access_spider', 'T.O.P._zh.wikipedia.org_all-access_spider', 'TANet_zh.wikipedia.org_all-access_spider']","['Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views', 'Page Views']",CC BY-SA 3.0,Web Traffic Time Series Forecasting,"A definitive benchmark for web server load forecasting, tracking the daily hits of approximately 145,000 specific Wikipedia articles. The dataset is characterized by noisy, bursty data distributions driven by real-world cultural trends, distinguishing between human users and automated web-spiders across various language locales.","Maggie, Oren Anava, Vitaly Kuznetsov, and Will Cukierski. (2017). Web Traffic Time Series Forecasting [Data set]. Kaggle.",https://www.kaggle.com/c/web-traffic-time-series-forecasting/data,"Load train_1.csv (wide format: Page column + daily date columns). Select first N_PAGES=145 page rows (deterministic slice of raw file, not random). Reshape without melt/pivot: transpose so dates become rows and page names become columns. Parse transposed index as UTC daily timestamps; 145 variates named by Wikipedia page identifiers. Frequency check: D (daily). Fill missing: standard daily reindex on all 145 page series. Transform: 145 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Web_Daily),wikipedia_web_multivariate_covariate,Web_Daily,covariate_web_daily_wikipedia_web_traffic,['Python_zh.wikipedia.org_all-access_spider'],['Page Views'],CC BY-SA 3.0,Web Traffic Time Series Forecasting,"A definitive benchmark for web server load forecasting, tracking the daily hits of approximately 145,000 specific Wikipedia articles. The dataset is characterized by noisy, bursty data distributions driven by real-world cultural trends, distinguishing between human users and automated web-spiders across various language locales.","Maggie, Oren Anava, Vitaly Kuznetsov, and Will Cukierski. (2017). Web Traffic Time Series Forecasting [Data set]. Kaggle.",https://www.kaggle.com/c/web-traffic-time-series-forecasting/data,"Load train_1.csv (wide format: Page column + daily date columns). Select first N_PAGES=145 page rows (deterministic slice of raw file, not random). Reshape without melt/pivot: transpose so dates become rows and page names become columns. Parse transposed index as UTC daily timestamps; 145 variates named by Wikipedia page identifiers. Frequency check: D (daily). Fill missing: standard daily reindex on all 145 page series. Transform: 145 variates; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Econ._Quarterly),us_gdp_econ_multivariate,Econ._Quarterly,multivariate_econ_quarterly_fred_qd,"['GDPC1', 'PCECC96', 'PCDGx', 'PCESVx', 'PCNDx', 'GPDIC1', 'FPIx', 'Y033RC1Q027SBEAx', 'PNFIx', 'PRFIx', 'A014RE1Q156NBEA', 'GCEC1', 'A823RL1Q225SBEA', 'FGRECPTx', 'SLCEx', 'EXPGSC1', 'IMPGSC1', 'DPIC96', 'OUTNFB', 'OUTBS', 'OUTMS', 'INDPRO', 'IPFINAL', 'IPCONGD', 'IPMAT', 'IPDMAT', 'IPNMAT', 'IPDCONGD', 'IPB51110SQ', 'IPNCONGD', 'IPBUSEQ', 'IPB51220SQ', 'TCU', 'CUMFNS', 'PAYEMS', 'USPRIV', 'MANEMP', 'SRVPRD', 'USGOOD', 'DMANEMP', 'NDMANEMP', 'USCONS', 'USEHS', 'USFIRE', 'USINFO', 'USPBS', 'USLAH', 'USSERV', 'USMINE', 'USTPU', 'USGOVT', 'USTRADE', 'USWTRADE', 'CES9091000001', 'CES9092000001', 'CES9093000001', 'CE16OV', 'CIVPART', 'UNRATE', 'UNRATESTx', 'UNRATELTx', 'LNS14000012', 'LNS14000025', 'LNS14000026', 'UEMPLT5', 'UEMP5TO14', 'UEMP15T26', 'UEMP27OV', 'LNS13023621', 'LNS13023557', 'LNS13023705', 'LNS13023569', 'LNS12032194', 'HOABS', 'HOAMS', 'HOANBS', 'AWHMAN', 'AWHNONAG', 'AWOTMAN', 'HWIx', 'HOUST', 'HOUST5F', 'PERMIT', 'HOUSTMW', 'HOUSTNE', 'HOUSTS', 'HOUSTW', 'CMRMTSPLx', 'RSAFSx', 'AMDMNOx', 'ACOGNOx', 'AMDMUOx', 'ANDENOx', 'INVCQRMTSPL', 'PCECTPI', 'PCEPILFE', 'GDPCTPI', 'GPDICTPI', 'IPDBS', 'DGDSRG3Q086SBEA', 'DDURRG3Q086SBEA', 'DSERRG3Q086SBEA', 'DNDGRG3Q086SBEA', 'DHCERG3Q086SBEA', 'DMOTRG3Q086SBEA', 'DFDHRG3Q086SBEA', 'DREQRG3Q086SBEA', 'DODGRG3Q086SBEA', 'DFXARG3Q086SBEA', 'DCLORG3Q086SBEA', 'DGOERG3Q086SBEA', 'DONGRG3Q086SBEA', 'DHUTRG3Q086SBEA', 'DHLCRG3Q086SBEA', 'DTRSRG3Q086SBEA', 'DRCARG3Q086SBEA', 'DFSARG3Q086SBEA', 'DIFSRG3Q086SBEA', 'DOTSRG3Q086SBEA', 'CPIAUCSL', 'CPILFESL', 'WPSFD49207', 'PPIACO', 'WPSFD49502', 'WPSFD4111', 'PPIIDC', 'WPSID61', 'WPU0531', 'WPU0561', 'OILPRICEx', 'AHETPIx', 'CES2000000008x', 'CES3000000008x', 'COMPRMS', 'COMPRNFB', 'RCPHBS', 'OPHMFG', 'OPHNFB', 'OPHPBS', 'ULCBS', 'ULCMFG', 'ULCNFB', 'UNLPNBS', 'FEDFUNDS', 'TB3MS', 'TB6MS', 'GS1', 'GS10', 'MORTGAGE30US', 'AAA', 'BAA', 'BAA10YM', 'MORTG10YRx', 'TB6M3Mx', 'GS1TB3Mx', 'GS10TB3Mx', 'CPF3MTB3Mx', 'BOGMBASEREALx', 'M1REAL', 'M2REAL', 'BUSLOANSx', 'CONSUMERx', 'NONREVSLx', 'REALLNx', 'REVOLSLx', 'TOTALSLx', 'DRIWCIL', 'TABSHNOx', 'TLBSHNOx', 'LIABPIx', 'TNWBSHNOx', 'NWPIx', 'TARESAx', 'HNOREMQ027Sx', 'TFAABSHNOx', 'VIXCLSx', 'USSTHPI', 'SPCS10RSA', 'SPCS20RSA', 'TWEXAFEGSMTHx', 'EXUSEU', 'EXSZUSx', 'EXJPUSx', 'EXUSUKx', 'EXCAUSx', 'UMCSENTx', 'USEPUINDXM', 'B020RE1Q156NBEA', 'B021RE1Q156NBEA', 'GFDEGDQ188S', 'GFDEBTNx', 'IPMANSICS', 'IPB51222S', 'IPFUELS', 'UEMPMEAN', 'CES0600000007', 'TOTRESNS', 'NONBORRES', 'GS5', 'TB3SMFFM', 'T5YFFM', 'AAAFFM', 'WPSID62', 'PPICMM', 'CPIAPPSL', 'CPITRNSL', 'CPIMEDSL', 'CUSR0000SAC', 'CUSR0000SAD', 'CUSR0000SAS', 'CPIULFSL', 'CUSR0000SA0L2', 'CUSR0000SA0L5', 'CES0600000008', 'DTCOLNVHFNM', 'DTCTHFNM', 'INVEST', 'HWIURATIOx', 'CLAIMSx', 'BUSINVx', 'ISRATIOx', 'CONSPIx', 'CP3M', 'COMPAPFF', 'PERMITNE', 'PERMITMW', 'PERMITS', 'PERMITW', 'NIKKEI225', 'NASDAQCOM', 'CUSR0000SEHC', 'TLBSNNCBx', 'TLBSNNCBBDIx', 'TTAABSNNCBx', 'TNWMVBSNNCBx', 'TNWMVBSNNCBBDIx', 'TLBSNNBx', 'TLBSNNBBDIx', 'TABSNNBx', 'TNWBSNNBx', 'TNWBSNNBBDIx', 'CNCFx', 'S&P 500', 'S&P div yield', 'S&P PE ratio']","['Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)', 'Macro Unit (Index/Rate/USD/Persons)']",Public Domain,FRED-QD: A Quarterly Database for Macroeconomic Research,"Engineered by the Federal Reserve Bank of St. Louis, this massive econometric matrix harmonizes over 200 distinct quarterly time series into a single dense panel. It categorizes macro indicators into 14 groups—including real GDP, personal consumption, industrial production, labor markets, price indices, and interest rates—making it the definitive benchmark for training and evaluating large-scale ""big data"" forecasting algorithms.","McCracken, M. W., & Ng, S. (2020). FRED-QD: A Quarterly Database for Macroeconomic Research. Federal Reserve Bank of St. Louis.",https://research.stlouisfed.org/econ/mccracken/fred-databases/,"Load 2026-04-QD.csv. Drop first two rows (factors and transform metadata rows in sasdate column); data begins at 1959-Q1. Rename sasdate to timestamps; parse as UTC datetime. Kept all 245 FRED series columns (every column except sasdate); no series removed. Examples: GDPC1, PCECC96, INDPRO, CPIAUCSL, FEDFUNDS, UNRATE, S&P 500, etc. Frequency check: quarterly grid (QS or median-interval fallback for ~90-day spacing). Fill missing: standard quarterly reindex; early missing values per series retained as NaN. Transform: 245 variates; variable_name = raw FRED series id; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Econ._Quarterly),us_gdp_econ_multivariate_covariate,Econ._Quarterly,covariate_econ_quarterly_fred_qd,['GDPC1'],['Macro Unit (Index/Rate/USD/Persons)'],Public Domain,FRED-QD: A Quarterly Database for Macroeconomic Research,"Engineered by the Federal Reserve Bank of St. Louis, this massive econometric matrix harmonizes over 200 distinct quarterly time series into a single dense panel. It categorizes macro indicators into 14 groups—including real GDP, personal consumption, industrial production, labor markets, price indices, and interest rates—making it the definitive benchmark for training and evaluating large-scale ""big data"" forecasting algorithms.","McCracken, M. W., & Ng, S. (2020). FRED-QD: A Quarterly Database for Macroeconomic Research. Federal Reserve Bank of St. Louis.",https://research.stlouisfed.org/econ/mccracken/fred-databases/,"Load 2026-04-QD.csv. Drop first two rows (factors and transform metadata rows in sasdate column); data begins at 1959-Q1. Rename sasdate to timestamps; parse as UTC datetime. Kept all 245 FRED series columns (every column except sasdate); no series removed. Examples: GDPC1, PCECC96, INDPRO, CPIAUCSL, FEDFUNDS, UNRATE, S&P 500, etc. Frequency check: quarterly grid (QS or median-interval fallback for ~90-day spacing). Fill missing: standard quarterly reindex; early missing values per series retained as NaN. Transform: 245 variates; variable_name = raw FRED series id; variable_unit blank. Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Multivariate(Nature_Annually),carbon_nature_multivariate,Nature_Annually,multivariate_nature_annually_global_carbon_budget,"['population', 'gdp', 'co2', 'co2_per_capita', 'co2_per_gdp', 'co2_growth_abs', 'co2_growth_prct', 'coal_co2', 'oil_co2', 'gas_co2', 'cement_co2', 'flaring_co2', 'land_use_change_co2', 'co2_including_luc', 'co2_including_luc_per_capita', 'cumulative_co2', 'cumulative_co2_including_luc', 'methane', 'nitrous_oxide', 'total_ghg', 'primary_energy_consumption', 'energy_per_capita', 'share_global_co2', 'share_global_cumulative_co2', 'trade_co2', 'temperature_change_from_co2']","['Count', 'International Dollars', 'Million Tonnes', 'Tonnes per capita', 'kg per USD', 'Million Tonnes', 'Percentage', 'Million Tonnes', 'Million Tonnes', 'Million Tonnes', 'Million Tonnes', 'Million Tonnes', 'Million Tonnes', 'Million Tonnes', 'Tonnes per capita', 'Million Tonnes', 'Million Tonnes', 'Million Tonnes CO2e', 'Million Tonnes CO2e', 'Million Tonnes CO2e', 'TWh', 'kWh', 'Percentage', 'Percentage', 'Million Tonnes', 'Celsius']",CC BY 4.0,Global CO2 Emissions 1750-2024,"A foundational dataset mapping the historical anthropogenic footprint on the natural world since the start of the Industrial Revolution. It contextualizes annual terrestrial CO2 emissions and atmospheric temperature changes by aligning them against human population growth, GDP, specialized fuel consumptions (coal, oil, gas, cement), and global land-use alterations.","Global Carbon Project, Our World in Data (2024). Global CO2 Emissions 1750-2024 [Data set]. Kaggle.",https://www.kaggle.com/datasets/elvisbui/global-co2-emissions-1750-2024,"Load global_co2_emissions_1750_2024.csv. Select one country with np.random.default_rng(seed=42) → Barbados. Filter and sort by year; rename year to timestamps (UTC). Removed identifier columns from variates: country, year, iso_code. Kept 26 numeric variates: population, gdp, co2, co2_per_capita, co2_per_gdp, co2_growth_abs, co2_growth_prct, coal_co2, oil_co2, gas_co2, cement_co2, flaring_co2, land_use_change_co2, co2_including_luc, co2_including_luc_per_capita, cumulative_co2, cumulative_co2_including_luc, methane, nitrous_oxide, total_ghg, primary_energy_consumption, energy_per_capita, share_global_co2, share_global_cumulative_co2, trade_co2, temperature_change_from_co2. Frequency check: annual (YS or ~365-day median fallback). Fill missing: standard annual reindex. Transform: 26 variates for Barbados; variable_unit blank. Output file: multivariate_nature_annually_global_carbon_budget.csv (Barbados subset). Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."
Covariate(Nature_Annually),carbon_nature_multivariate_covariate,Nature_Annually,covariate_nature_annually_global_carbon_budget,['co2'],['Million Tonnes'],CC BY 4.0,Global CO2 Emissions 1750-2024,"A foundational dataset mapping the historical anthropogenic footprint on the natural world since the start of the Industrial Revolution. It contextualizes annual terrestrial CO2 emissions and atmospheric temperature changes by aligning them against human population growth, GDP, specialized fuel consumptions (coal, oil, gas, cement), and global land-use alterations.","Global Carbon Project, Our World in Data (2024). Global CO2 Emissions 1750-2024 [Data set]. Kaggle.",https://www.kaggle.com/datasets/elvisbui/global-co2-emissions-1750-2024,"Load global_co2_emissions_1750_2024.csv. Select one country with np.random.default_rng(seed=42) → Barbados. Filter and sort by year; rename year to timestamps (UTC). Removed identifier columns from variates: country, year, iso_code. Kept 26 numeric variates: population, gdp, co2, co2_per_capita, co2_per_gdp, co2_growth_abs, co2_growth_prct, coal_co2, oil_co2, gas_co2, cement_co2, flaring_co2, land_use_change_co2, co2_including_luc, co2_including_luc_per_capita, cumulative_co2, cumulative_co2_including_luc, methane, nitrous_oxide, total_ghg, primary_energy_consumption, energy_per_capita, share_global_co2, share_global_cumulative_co2, trade_co2, temperature_change_from_co2. Frequency check: annual (YS or ~365-day median fallback). Fill missing: standard annual reindex. Transform: 26 variates for Barbados; variable_unit blank. Output file: multivariate_nature_annually_global_carbon_budget.csv (Barbados subset). Output CSV schema: variable_name, variable_unit, timestamps (JSON array of ISO-8601 UTC strings), values (JSON array; NaN serialized as null via allow_nan=True). Timestamps are shared across variates in multivariate tasks (one row per variate)."