Skip to contents

Compares eta estimates from a validation run against estimates from the full data run. Calculates error metrics (median error, MAE) and coverage statistics to assess how well the validation estimates match the full data estimates.

Usage

summarize_eta_all(res_val, res_all, year_select = 2023)

Arguments

res_val

Tibble with columns: draw, eta, iso, year from validation run

res_all

Tibble with columns: draw, eta, iso, year from full data run

year_select

Year to summarize (default: 2023)

Value

A tibble with summary statistics:

ncountries

Number of countries

ME

Median error (full data median - validation median)

MAE

Median absolute error

MeanE

Mean error

MeanAE

Mean absolute error

Coverage

Proportion of full data draws within validation 95% CI

Prop_above_CI

Proportion of full data draws above validation CI

Prop_below_CI

Proportion of full data draws below validation CI