Dark Mode
Prophet-Ready COVID-19 Holiday Dynamics
Patient Health Records & Digital Health
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This dataset is designed to assist in the analysis of the impact of holidays on confirmed coronavirus disease statistics. It facilitates the study of how holidays affect the accuracy of predicting coronavirus cases, aiding in the identification of new patterns and the formulation of optimal strategies to counteract or minimise disease spread. The data is structured for use with the Prophet forecasting tool, incorporating holiday information for various countries and their regions. It addresses limitations of existing holiday libraries by including dates that have been adjusted to account for the typical delay between infection and symptom manifestation, typically by shifting holiday dates forward by seven days to align with when a person might enter statistics. The dataset contains refined data and includes additional countries beyond initial versions of holiday libraries.
Columns
- ds_holidays: The real date of the holiday, in DateTime format.
- holiday: The name of the holiday, as a String.
- ds: The shifted date of the holiday, in DateTime format. This date is typically moved forward by 7 days from the real holiday date to account for symptom manifestation delays.
- country: The name of the country, using its common name from ISO 3166 where available, otherwise its name from ISO 3166. This is a String.
- code: The two-letter country code (alpha_2) according to ISO 3166, as a String.
- country_official_name: The official name of the country, using its official name from ISO 3166 where available, otherwise its name from ISO 3166. This is a String.
- lower_window: The lower boundary of the holiday impact window, a parameter for the Prophet model, measured in days. The default value is -3.
- upper_window: The upper boundary of the holiday impact window, a parameter for the Prophet model, measured in days. The default value is 3.
- prior_scale: The prior scale parameter for the holiday component in the Prophet model. The default value is 10.
- source: The URL or string indicating the source of the real holiday date and name.
Distribution
The dataset is provided in a CSV format, specifically as
holidays_df_of_67_countries_for_covid_19.csv. The file size is approximately 101.08 kB and it contains 10 columns. The dataset includes 760 valid records as of the provided sample.Usage
This dataset is ideal for:
- Studying the influence of public holidays on the accuracy of COVID-19 disease predictions.
- Identifying novel patterns related to coronavirus spread during holiday periods.
- Developing and testing optimal solutions to mitigate or minimise the spread of coronavirus.
- Forecasting confirmed coronavirus cases using Prophet technology, specifically by incorporating holiday effects.
- Modelling the correlation between holidays and the number of coronavirus patients.
Coverage
The dataset covers 70 countries, which include 62 countries available as of June 2020, expanded to 67 by January 2021 with the addition of Burundi (BI), Latvia (LV), Morocco (MA), Romania (RO), and Vietnam (VN). Further additions include Albania (AL), Georgia (GE), and Moldova (MD). The temporal scope for holiday dates spans from 20 January 2020 to 31 December 2021. The shifted holiday dates extend to 7 January 2021. While some countries may include regional data, this dataset does not specifically detail holiday impacts for individual regions within countries.
License
Attribution 4.0 International (CC BY 4.0)
Who Can Use It
This dataset is suitable for:
- Data scientists and modellers: To enhance time series forecasting models, especially those using Prophet, by accounting for holiday seasonality in disease statistics.
- Public health researchers and epidemiologists: To understand the societal factors influencing disease transmission dynamics during festive periods.
- Government agencies and policymakers: To inform public health interventions and policy decisions related to holiday gatherings and travel.
- Academic institutions: For research into the socio-behavioural aspects of pandemic spread and the development of predictive models.
Dataset Name Suggestions
- COVID-19 Holiday Impact Forecasting Data
- Global Coronavirus Holiday Statistics
- Prophet-Ready COVID-19 Holiday Dynamics
- Public Holidays and COVID-19 Cases Dataset
- Pandemic Holiday Influence Data
Attributes
Original Data Source: Prophet-Ready COVID-19 Holiday Dynamics
QUALITY