Process GTFCC Vaccination Data for MOSAIC Model

This function processes raw GTFCC vaccination data that has been scraped from the GTFCC OCV dashboard by the ees-cholera-mapping repository. It transforms the event-based data structure into a clean, structured dataset matching the WHO vaccination data format for use in the MOSAIC model.

Usage

process_GTFCC_vaccination_data(PATHS)

Arguments

PATHS

A list containing file paths for input and output data. The list should include:

• MODEL_INPUT: Path to the directory where processed vaccination data will be saved.

Value

The function saves the processed vaccination data to a CSV file in the directory specified by PATHS$MODEL_INPUT. It also returns the processed data as a data frame for further use in R.

Details

Each output row is one GTFCC request (req_id, e.g. "2019-I05-D01"): doses_shipped is the sum of all its Delivery events and campaign_date the first delivery date. Four columns carry the request's delivery and round structure (MOSAIC v0.103.0); the round columns are read from its Round events (round_id C<campaign>-R<round>, doses = doses administered in that round):

req_id

The GTFCC request identifier, for tracing a row back to the raw log.

delivery_schedule

The request's deliveries as <date>:<doses> pairs separated by ;, in date order, with deliveries on the same date summed, e.g. "2019-04-24:835200;2019-09-25:835200". Their doses sum to doses_shipped; est_vaccination_rate releases each delivery on its own date.

round_sequence

The request's doses as an ordered sequence of <dose>:<weight> blocks separated by ;, e.g. "1:357200;2:332900;1:562000;2:693900". <dose> is 1 for a first round (R01) and 2 for a second or later round (R02+), the blocks run in campaign order with R01 before R02 inside a campaign (campaign numbers follow administration order), consecutive rounds of the same dose are merged, and <weight> is the doses administered in the block (a campaign-round listed more than once counts once, with its reported doses). Weights are relative: est_vaccination_rate splits the shipped doses across the blocks in proportion to them, so a single block is written with weight 1. NA when the round is unknown.

round_basis

How the sequence was obtained: "rounds" (every round reports its doses, or all rounds are of one dose so the counts are not needed), "rounds_imputed" (the request has first and second rounds but some round dose counts are unreported; each unreported round carries the mean reported round of the request, or all rounds weigh equally when none is reported) or "unknown" (no Round event with a C##-R## identifier: the doses count as first doses downstream).

This function performs the following steps:

1. Load and Transform GTFCC Data: - Reads the scraped GTFCC data from ees-cholera-mapping repository. - Transforms event-based structure (Request, Decision, Delivery, Round events) to request-based structure.

2. Map to WHO Format: - Aggregates events by request ID to extract doses requested, approved, and shipped. - Uses Delivery event dates as campaign dates. - Records each delivery's date and doses, and attributes the request's doses to first and second rounds from its Round events. - Converts country names to ISO codes for consistency.

3. Infer Missing Campaign Dates: - Computes the delay between decision and campaign dates. - Infers missing campaign dates based on the mean delay where possible. - Fixes missing decision dates for grouped request numbers.

4. Validate and Filter Data: - Ensures all rows have valid campaign dates. - Removes rows where doses_shipped is zero or missing. - Filters for MOSAIC countries only.

5. Save Processed Data: - Writes the cleaned and processed dataset to a CSV file for further modeling.

Examples

if (FALSE) { # \dontrun{
# Example usage
PATHS <- list(
  MODEL_INPUT = "path/to/model/input"
)

processed_data <- process_GTFCC_vaccination_data(PATHS)
} # }