target-data

Target Data

This folder is used to store target data (also called "ground truth" or "truth data") relevant to the RSV modeling efforts. It contains the data and code used to generate the data.

RSV-NET data

The Weekly Rates of Laboratory-Confirmed RSV Hospitalizations from the RSV-NET Surveillance System is used for the hospitalization target after standardization. The standardization process is available in the get_target_data.R script

The source data is downloaded and updated weekly in the auxiliary-data/ folder. Please consult, the associated documentation for more information.

Following the source data update, the target data target_data/DATE_rsvnet_hospitalization.csv is updated weekly with update target GitHub Action.

Workflow

As the source format of the data change on the 2023-11-09, the workflow and R script have been updated to adapt to the changes:

• the overall age group has been removed from the selection for the seasons: 2014-2015, 2015-2016, 2016-2017, 2017-2018
• the two seasons 2014-2015 and 2015-2016 has been removed
• some value and variable have been renamed, for example "All" instead of "Overall"

The RSV-NET source files is standardized following these steps:

1. Load the RSV-NET file from the auxiliary-data/ folder

2. Filter the source files to keep only the information of interest:

   • Select overall race and sex and age groups of interest (hub standard format in parentheses) :

     • 0-4 years ("0-4"), 5-17 years ("5-17"), 18-49 years ("18-49"), 50-64 years ("50-64"), 65+ years ("65-130"), 0-<6 months ("0-0.49"), 6-<12 months ("0.5-0.99"), 1-<2 years ("1-1.99"), 2-4 years ("2-4"), 18+ (Adults) ("18-130") and overall ("0-130")

     • Deprecated step as the overall value for these seasons is not included in the data anymore As the data before the 2018-2019 season does not include children data, the overall age group has been removed from the selection for the seasons: 2014-2015, 2015-2016, 2016-2017, 2017-2018. For more information, please consult the RSV-NET Overview and Methods webpage.

   • Remove the seasonal summaries

3. Re-code variable and associated values to the hub standard. See model-output/README.md for more information, including adding all the missing values to obtain a "square" time series

4. Calculate the hospitalization number by applying:

   • rate * population size / 100000
   • For the 6 months age group, the population size for the corresponding year divided by 2 is used.
   • The population data from the year 2022 are used for the year 2023
   • The population size information comes from the US Census Bureau, please consult the auxiliary-data/ for more information

5. Standardize the output to the hub format:

   • output with six columns:
     • location: fips code of the associated state, "US" for national level
     • date: end of the epiweek in a "YYYY-MM-DD" format
     • age_group: age groups in the hub format
     • target: associated target:
       • "inc hosp": number of hospitalization (target variable)
       • "rate hosp": number of hospitalization per 100 000
     • value: associated value
     • population: associated population size

6. Append previous season: 2014-2015 and 2015-2016 (removed on the new version of the data (2023-11-10)) by using historical data containing the last version of the 2014-2015 and 2015-2016 seasons (commit #f183e8a)

7. Archive past version of the output

8. Write the output in a CSV format with the date in the filename

Past version

Past weekly versions of the file are available in the archive/ folder.

Visualization

A function to visualized the target data is also available in this folder.

The function weekly_hosp_state() is available in the target_data_viz.R script.

The function can take as input the target data filtered for a specific target and returns a faceted plot with the target by location for a specific age group. For more information on the function, please consult the documentation in the target_data_viz.R script.