41. R¶

Use httr to request a NetCDF file and ncdf4 to inspect it. The example uses IMERG V07 daily satellite rainfall.

41.1 Install and authenticate¶

install.packages(c("httr", "ncdf4"))

Set GAP_API_TOKEN in your environment before starting R or RStudio. The sample reads Sys.getenv("GAP_API_TOKEN"); it does not contain a key. See API keys.

41.2 Download and run¶

Download sample.R and run it in RStudio or with Rscript sample.R.

The request uses:

params <- list(
  product = "imerg_v07",
  attributes = "precipitation",
  start_date = "2026-09-01", end_date = "2026-09-03",
  output_type = "netcdf", bbox = "36.7,-1.4,36.9,-1.2"
)

The script checks HTTP status, validates the NetCDF file, prints the variables and rainfall range, and saves data.nc. It closes the NetCDF handle and removes the temporary file even if processing fails.

41.3 Multiple points¶

Download sample_points.R. Provide points.csv with numeric lon and lat columns in WGS84 decimal degrees. It downloads one CSV per point and stops on an HTTP error. Reduce request frequency if the API responds with 429.

41.4 Change products¶

Choose field identifiers from the product reference. For a forecast, update the dates to an available run and valid period. NextGen daily rainfall uses names such as precip_accum_p50; historical CBAM uses a different field vocabulary.

Do not assume NetCDF dimension order. Inspect nc$var and coordinates before extracting or plotting a slice. Missing values are not zero. A forecast percentile is not the percentile of a sum across days.