Extract Under-5 Mortality Values from IHME Raster Stack

This function extracts under-5 mortality values from IHME (Institute for Health Metrics and Evaluation) raster data for specified geographic areas using exact extraction methods. The data source is the IHME's geospatial estimates for under-5 mortality rates in low and middle-income countries (LMICs) from 2000-2017.

Usage

process_ihme_u5m_raster(
  shape,
  raster_stack,
  id_col = c("adm1", "adm2"),
  rates = TRUE,
  stat_type = "mean"
)

Source

https://ghdx.healthdata.org/record/ihme-data/lmic-under5-mortality-rate-geospatial-estimates-2000-2017

Arguments

shape

An sf object containing the geographic areas for extraction

raster_stack

A terra SpatRaster or raster stack containing IHME mortality data

id_col

Character vector of column names in shape to use as identifiers (default: c("adm1", "adm2"))

rates

Logical indicating whether values are rates (TRUE) or counts (TRUE, default)

stat_type

Character string specifying statistic type: "mean", "median", or "both". Only applies when rates = TRUE. For counts, sum is always used.

Value

A tibble containing extracted values for each area and year with columns for area identifiers, year, and mortality values. For rates, values represent deaths per 1,000 live births. For counts, values represent total number of deaths.

Details

The function processes each layer in the IHME raster stack, which represents annual estimates starting from 2000. For rates, mean or median values are extracted based on stat_type; for counts, sums are used. The IHME data provides high-resolution (5 x 5 km) estimates of under-5 mortality across LMICs.

Note

The IHME raster data represents model-based estimates and should be usedwith consideration of the underlying methodology and uncertainties described in the source documentation.

Examples

if (FALSE) { # \dontrun{
# Load required libraries and data
library(sf)
library(terra)

# Example with rates using mean (default)
mortality_rates_mean <- process_ihme_u5m_raster(
  shape = admin_boundaries,
  raster_stack = mortality_raster,
  id_col = c("adm1", "adm2"),
  rates = TRUE,
  stat_type = "mean"
)

# Example with rates using median
mortality_rates_median <- process_ihme_u5m_raster(
  shape = admin_boundaries,
  raster_stack = mortality_raster,
  id_col = c("adm1", "adm2"),
  rates = TRUE,
  stat_type = "median"
)

# Example with both mean and median
mortality_rates_both <- process_ihme_u5m_raster(
  shape = admin_boundaries,
  raster_stack = mortality_raster,
  id_col = c("adm1", "adm2"),
  rates = TRUE,
  stat_type = "both"
)

# Example with counts (total deaths - always uses sum)
mortality_counts <- process_ihme_u5m_raster(
  shape = admin_boundaries,
  raster_stack = mortality_raster,
  id_col = c("adm1"),
  rates = FALSE
)
} # }