Context

This research is devoted to the analysis of the impact of holidays on the statistics of confirmed coronavirus diseases. The Prophet using the holidays library with holidays of countries and their regions. As of 30 June 2020, only 62 countries (some with regions) are available in the holidays library:

['AR', 'AT', 'AU', 'BD', 'BE', 'BG', 'BR', 'BY', 'CA', 'CH', 'CL', 'CN', 'CO', 'CZ', 'DE', 'DK', 'DO', 'EE', 'EG', 'ES', 'FI', 'FR', 'GB', 'GR', 'HN', 'HR', 'HU', 'ID', 'IE', 'IL', 'IN', 'IS', 'IT', 'JP', 'KE', 'KR', 'LT', 'LU', 'MX', 'MY', 'NG', 'NI', 'NL', 'NO', 'NZ', 'PE', 'PH', 'PK', 'PL', 'PT', 'PY', 'RS', 'RU', 'SE', 'SG', 'SI', 'SK', 'TH', 'TR', 'UA', 'US', 'ZA'] or ['Argentina', 'Australia', 'Austria', 'Bangladesh', 'Belarus', 'Belgium', 'Brazil', 'Bulgaria', 'Canada', 'Chile', 'China', 'Colombia', 'Croatia', 'Czechia', 'Denmark', 'Dominican Republic', 'Egypt', 'Estonia', 'Finland', 'France', 'Germany', 'Greece', 'Honduras', 'Hungary', 'Iceland', 'India', 'Indonesia', 'Ireland', 'Israel', 'Italy', 'Japan', 'Kenya', 'Korea, Republic of', 'Lithuania', 'Luxembourg', 'Malaysia', 'Mexico', 'Netherlands', 'New Zealand', 'Nicaragua', 'Nigeria', 'Norway', 'Pakistan', 'Paraguay', 'Peru', 'Philippines', 'Poland', 'Portugal', 'Russian Federation', 'Serbia', 'Singapore', 'Slovakia', 'Slovenia', 'South Africa', 'Spain', 'Sweden', 'Switzerland', 'Thailand', 'Turkey', 'Ukraine', 'United Kingdom', 'United States']

I will note at once that the list of available countries in the description of the holidays library contains a lot of mistakes, which I wrote to the authors.

When I asked if this list would expand, the Prophet team made it clear that they were waiting for help from the community with holidays library expand.

As of Jan 2021 (version 8.4.1), 67 countries (some with regions) are available in the holidays library: a number of data have been refined and countries ['BI', 'LV', 'MA', 'RO', 'VN' - two-letter country codes or alpha_2 of the country (ISO 3166)] added.

Unfortunately, the format of the holidays library is not very suitable for coronavirus problems, as it has a number of disadvantages. First, the names of the countries are given in one word, which makes it difficult for many of them to identify them according to their common names (ISO 3166). It is best that the dataset contains the common name and two-letter abbreviation in English according to ISO 3166 (see pycountry). Second, the dates are not adapted to the potential impact of the holidays on coronavirus statistics. It is known that after the moment of infection, the active manifestation of symptoms occurs with a delay of 4-10 days, that is a person is likely to get into the statistics on the number of diseases only after 4-7 days. Therefore, it is advisable to use the dates window of impacts: ``` Lower_window = [4, 7] Upper_window = [7, 10]

`Lower_window <= 0`
But my [request](https://github.com/facebook/prophet/issues/1588#issue-661098613) to allow positive numbers in this parameter [was refused](https://github.com/facebook/prophet/issues/1588#issuecomment-661984730) by the Prophet team and [advised](https://github.com/facebook/prophet/issues/1588#issuecomment-661984730) to simply move the dates themselves.
Therefore, it is advisable to shift the holiday dates by 7 days. If the researcher thinks that 7 is too much and enough is 4 days, then he simply indicates "Lower" of the window in -3. Actually, by default, it makes sense to specify parameters:

Lower_window = -3 Upper_window = 3

If necessary, these settings are easy to change

### Content

This dataset:
1. Contains ISO codes, ISO names (common and official) (ISO 3166) of **70** countries (3 European countries **['Albania' - 'AL', 'Georgia' - 'GE', 'Moldova' - 'MD']** have been added).
2. Contains imported dates from the holidays library for 2020-01-20-2021-12-31 (all countries from holidays library as of Jan 2021), and the same dates, but moved 7 days forward.
3. Holidays of countries that are not in the list of holidays of the library, but which are in the data of the World Health Organization and on which considerable statistics of diseases on coronavirus are already collected.
4. Parameters for Prophet model:
`lower_window, upper_window, prior_scale`
If you find errors, please write to the [Discussion](https://www.kaggle.com/vbmokin/covid19-holidays-of-countries/discussion).

It is planned to periodically update (and, if necessary, correct) this dataset. 

### Acknowledgements

Thanks to the authors of the...

The ongoing coronavirus pandemic has strongly impacted the shopping behavior of consumers in the United States and recent survey data indicates that consumers do not feel that this situation will be resolved in the upcoming holiday season. A May 2020 survey of U.S. consumers found that compared to last year, 49 percent of respondents were more interested in shopping online for the holidays. A third of respondents was also more interested in buying online and picking their order up in-store.

Note: After May 3, 2024, this dataset will no longer be updated because hospitals are no longer required to report data on COVID-19 hospital admissions, hospital capacity, or occupancy data to HHS through CDC’s National Healthcare Safety Network (NHSN). The related CDC COVID Data Tracker site was revised or retired on May 10, 2023.

This dataset represents weekly COVID-19 hospitalization data and metrics aggregated to national, state/territory, and regional levels. COVID-19 hospitalization data are reported to CDC’s National Healthcare Safety Network, which monitors national and local trends in healthcare system stress, capacity, and community disease levels for approximately 6,000 hospitals in the United States. Data reported by hospitals to NHSN and included in this dataset represent aggregated counts and include metrics capturing information specific to COVID-19 hospital admissions, and inpatient and ICU bed capacity occupancy.

Reporting information:

• As of December 15, 2022, COVID-19 hospital data are required to be reported to NHSN, which monitors national and local trends in healthcare system stress, capacity, and community disease levels for approximately 6,000 hospitals in the United States. Data reported by hospitals to NHSN represent aggregated counts and include metrics capturing information specific to hospital capacity, occupancy, hospitalizations, and admissions. Prior to December 15, 2022, hospitals reported data directly to the U.S. Department of Health and Human Services (HHS) or via a state submission for collection in the HHS Unified Hospital Data Surveillance System (UHDSS).
• While CDC reviews these data for errors and corrects those found, some reporting errors might still exist within the data. To minimize errors and inconsistencies in data reported, CDC removes outliers before calculating the metrics. CDC and partners work with reporters to correct these errors and update the data in subsequent weeks.
• Many hospital subtypes, including acute care and critical access hospitals, as well as Veterans Administration, Defense Health Agency, and Indian Health Service hospitals, are included in the metric calculations provided in this report. Psychiatric, rehabilitation, and religious non-medical hospital types are excluded from calculations.
• Data are aggregated and displayed for hospitals with the same Centers for Medicare and Medicaid Services (CMS) Certification Number (CCN), which are assigned by CMS to counties based on the CMS Provider of Services files.
• Full details on COVID-19 hospital data reporting guidance can be found here: https://www.hhs.gov/sites/default/files/covid-19-faqs-hospitals-hospital-laboratory-acute-care-facility-data-reporting.pdf

Metric details:

• Time Period: timeseries data will update weekly on Mondays as soon as they are reviewed and verified, usually before 8 pm ET. Updates will occur the following day when reporting coincides with a federal holiday. Note: Weekly updates might be delayed due to delays in reporting. All data are provisional. Because these provisional counts are subject to change, including updates to data reported previously, adjustments can occur. Data may be updated since original publication due to delays in reporting (to account for data received after a given Thursday publication) or data quality corrections.
• New COVID-19 Hospital Admissions (count): Number of new admissions of patients with laboratory-confirmed COVID-19 in the previous week (including both adult and pediatric admissions) in the entire jurisdiction.
• New COVID-19 Hospital Admissions (7-Day Average): 7-day average of new admissions of patients with laboratory-confirmed COVID-19 in the previous week (including both adult and pediatric admissions) in the entire jurisdiction.
• Cumulative COVID-19 Hospital Admissions: Cumulative total number of admissions of patients with laboratory-confirmed COVID-19 (including both adult and pediatric admissions) in the entire jurisdiction since August 1, 2020.
• Cumulative COVID-19 Hospital Admissions Rate: Cumulative total number of admissions of patients with laboratory-confirmed COVID-19 (including both adult and pediatric admissions) in the entire jurisdiction since August 1, 2020 divided by 2019 intercensal population estimate for that jurisdiction multiplied by 100,000.
• New COVID-19 Hospital Admissions Rate (7-day average) percent change from prior week: Percent change in the 7-day average new admissions of patients with laboratory-confirmed COVID-19 per 100,000 population compared with the prior week.
• New COVID-19 Hospital Admissions (7-Day Total): 7-day total number of new admissions of patients with laboratory-confirmed COVID-19 (including both adult and pediatric admissions) in the entire jurisdiction.
• New COVID-19 Hospital Admissions Rate (7-Day Total): 7-day total number of new admissions of patients with laboratory-confirmed COVID-19 (including both adult and pediatric admissions) for the entire jurisdiction divided by 2019 intercensal population estimate for that jurisdiction multiplied by 100,000.
• Total Hospitalized COVID-19 Patients: 7-day total number of patients currently hospitalized with laboratory-confirmed COVID-19 (including both adult and pediatric patients) for the entire jurisdiction.
• Total Hospitalized COVID-19 Patients (7-Day Average): 7-day average of the number of patients currently hospitalized with laboratory-confirmed COVID-19 (including both adult and pediatric patients) for the entire jurisdiction.
• COVID-19 Inpatient Bed Occupancy (7-Day Average): Percentage of all staffed inpatient beds occupied by patients with laboratory-confirmed COVID-19 (including both adult and pediatric patients) within the entire jurisdiction is calculated as an average of valid daily values within the past 7 days (e.g., if only three valid values, the average of those three is taken). Averages are separately calculated for the daily numerators (patients hospitalized with confirmed COVID-19) and denominators (staffed inpatient beds). The average percentage can then be taken as the ratio of these two values for the entire jurisdiction.
• COVID-19 Inpatient Bed Occupancy absolute change from prior week: The absolute change in the percent of staffed inpatient beds occupied by patients with laboratory-confirmed COVID-19 represents the week-over-week absolute difference between the 7-day average occupancy of patients with confirmed COVID-19 in staffed inpatient beds in the past 7 days, compared with the prior week, in the entire jurisdiction.
• COVID-19 ICU Bed Occupancy (7-Day Average): Percentage of all staffed inpatient beds occupied by adult patients with confirmed COVID-19 within the entire jurisdiction is calculated as a 7-day average of valid daily values within the past 7 days (e.g., if only three valid values, the average of those three is taken). Averages are separately calculated for the daily numerators (adult patients hospitalized with confirmed COVID-19) and denominators (staffed adult ICU beds). The average percentage can then be taken as the ratio of these two values for the entire jurisdiction.
• COVID-19 ICU Bed Occupancy absolute change from prior week: The absolute change in the percent of staffed ICU beds occupied by patients with laboratory-confirmed COVID-19 represents the week-over-week absolute difference between the average occupancy of patients with confirmed COVID-19 in staffed adult ICU beds for the past 7 days, compared with the prior week, in the in the entire jurisdiction.

Note: October 27, 2023: Due to a data processing error, reported values for avg_percent_inpatient_beds_occupied_covid_confirmed will appear lower than previously reported values by an average difference of less than 1%. Therefore, previously reported values for avg_percent_inpatient_beds_occupied_covid_confirmed may have been overestimated and should be interpreted with caution.

October 27, 2023: Due to a data processing error, reported values for abs_chg_avg_percent_inpatient_beds_occupied_covid_confirmed will differ from previously reported values by an average absolute difference of less than 1%. Therefore, previously reported values for abs_chg_avg_percent_inpatient_beds_occupied_covid_confirmed should be interpreted with caution.

December 29, 2023: Hospitalization data reported to CDC’s National Healthcare Safety Network (NHSN) through December 23, 2023, should be interpreted with caution due to potential reporting delays that are impacted by Christmas and New Years holidays. As a result, metrics including new hospital admissions for COVID-19 and influenza and hospital occupancy may be underestimated for the week ending December 23, 2023.

January 5, 2024: Hospitalization data reported to CDC’s National Healthcare Safety Network (NHSN) through December 30, 2023 should be interpreted with caution due to potential reporting delays that are impacted by Christmas and New Years holidays. As a result, metrics including new hospital admissions for COVID-19 and influenza and hospital occupancy may be underestimated for the week ending December 30, 2023.

This file contains workforce absence statistics for education settings from 12 October 2020 to 17 December 2020 and again following wider reopening of schools, from 8 March 2021 to 16 September 2021. It excludes half term terms (19th October - 23rd October, and 2nd November 2020), the national lockdown during the spring term (4 January to 5 March 2021), Easter data (29 March - 19 April 2021) and summer holiday (17 July 2021 - 6 September 2021). Data for workforce during the restricted opening of schools can be found in table 1e.Data is in this file has been scaled to account for non-response so it is nationally representative.

Topics extracted from entire tweets dataset.

Available on Journal website. (ZIP)

Bike-sharing trip frequency: Weekday-, weekend-, and bank holiday-specific fixed-effects regression results where the treatment is the first Covid-19 case.

In a survey conducted in the U.S. in August 2020, ** percent of respondents reported that the main reason they expected their holiday spending in 2020 to increase was because they were able to save more money during the coronavirus pandemic.

The total number of inbound holiday visits to the United Kingdom decreased sharply in the second quarter of 2020 over the previous quarter, due to the coronavirus (COVID-19) pandemic. Between April and June 2020, it was estimated that the UK recorded only about *** thousand holiday visits, as a result of the global health crisis and travel restrictions. In the second quarter of 2019, the UK reported roughly *** million international holiday visits overall.

Retail platforms have undergone an unprecedented global traffic increase between January 2019 and June 2020, surpassing even holiday season traffic peaks. Overall, retail websites generated almost ** billion visits in June 2020, up from ***** billion global visits in January 2020. This is of course due to the global coronavirus pandemic which has forced millions of people to stay at home in order to stop the spread of the virus. Due to many shelter at home orders and a desire to avoid crowded stores in places where it is possible to shop, consumers have turned to the internet to procure everyday items such as groceries or toilet paper.