Variance of household counts by region, calculated from niger_eas. Used for demonstrating Neyman optimal allocation in stratified sampling.
niger_eas_varianceA tibble with 8 rows and 2 columns:
Factor. Region name
Numeric. Variance of household counts within region
# View the variance data
niger_eas_variance
#> # A tibble: 8 × 2
#> region var
#> <fct> <dbl>
#> 1 Agadez 6249.
#> 2 Diffa 1950.
#> 3 Dosso 2771.
#> 4 Maradi 3207.
#> 5 Niamey 5728.
#> 6 Tahoua 2487.
#> 7 Tillabéri 2506.
#> 8 Zinder 2926.
# Neyman allocation minimizes variance for fixed sample size
sampling_design() |>
stratify_by(region, alloc = "neyman", variance = niger_eas_variance) |>
draw(n = 200) |>
execute(niger_eas, seed = 42)
#> == tbl_sample ==
#> Weights: 5.57 - 9.29 (mean: 7.68 )
#>
#> # A tibble: 200 × 11
#> region ea_id department strata hh_count pop_estimate .weight .sample_id
#> * <fct> <chr> <fct> <fct> <dbl> <dbl> <dbl> <int>
#> 1 Agadez Aga_04_0012 Tchirozér… Urban 431 2586 5.67 1
#> 2 Agadez Aga_03_0010 Bilma Urban 259 1554 5.67 2
#> 3 Agadez Aga_01_0001 Agadez Rural 59 413 5.67 3
#> 4 Agadez Aga_02_0012 Arlit Rural 137 959 5.67 4
#> 5 Agadez Aga_01_0010 Agadez Urban 192 1344 5.67 5
#> 6 Agadez Aga_03_0009 Bilma Rural 41 287 5.67 6
#> 7 Agadez Aga_02_0005 Arlit Urban 138 966 5.67 7
#> 8 Agadez Aga_02_0011 Arlit Rural 70 350 5.67 8
#> 9 Agadez Aga_01_0007 Agadez Rural 166 1162 5.67 9
#> 10 Diffa Dif_06_0004 Mainé-Sor… Rural 67 469 9.29 10
#> # ℹ 190 more rows
#> # ℹ 3 more variables: .stage <int>, .weight_1 <dbl>, .fpc_1 <int>