Further speedup bootstrap algo by 2x.

xKDR · Jan 22, 2023 · c60a95a · c60a95a
1 parent dff1ea1
commit c60a95a
Show file tree

Hide file tree

Showing 4 changed files with 32 additions and 34 deletions.
diff --git a/src/bootstrap.jl b/src/bootstrap.jl
@@ -31,13 +31,10 @@ function bootweights(design::SurveyDesign; replicates=4000, rng=MersenneTwister(
         psus = unique(cluster_sorted[!, design.cluster])
         npsus = [(count(==(i), cluster_sorted[!, design.cluster])) for i in psus]
         nh = length(psus)
-        randinds = rand(rng, 1:(nh), replicates, (nh-1))
+        cluster_weights = cluster_sorted[!, design.weights] 
         for replicate in 1:replicates
-            rh = zeros(Int, nh)
-            for i in randinds[replicate, :]
-                rh[i] += 1
-            end            
-            cluster_sorted[!, "replicate_" * string(replicate)] = vcat([fill(rh[i] * (nh / (nh-1)), npsus[i]) for i in 1:length(rh)]...) .* cluster_sorted[!, design.weights] 
+            randinds = rand(rng, 1:(nh), (nh-1))          
+            cluster_sorted[!, "replicate_" * string(replicate)] = vcat([fill((count(==(i), randinds)) * (nh / (nh-1)), npsus[i]) for i in 1:nh]...) .* cluster_weights
         end   
         push!(substrata_dfs, cluster_sorted)
     end

diff --git a/src/mean.jl b/src/mean.jl
@@ -13,15 +13,15 @@ julia> mean(:api00, clus_one_stage)
  Row │ mean     SE
      │ Float64  Float64
 ─────┼──────────────────
-   1 │ 644.169  23.2877
+   1 │ 644.169  23.4107
 
 julia> mean([:api00, :enroll], clus_one_stage)
 2×3 DataFrame
  Row │ names   mean     SE
      │ String  Float64  Float64
 ─────┼──────────────────────────
-   1 │ api00   644.169  23.2877
-   2 │ enroll  549.716  46.2597
+   1 │ api00   644.169  23.4107
+   2 │ enroll  549.716  45.7835
 ```
 """
 function mean(x::Symbol, design::ReplicateDesign)
@@ -52,17 +52,17 @@ julia> mean(:api00, :cname, clus_one_stage)
  Row │ cname        mean     SE
      │ String15     Float64  Any
 ─────┼───────────────────────────────────
-   1 │ Santa Clara  732.077  59.6794
-   2 │ San Diego    659.436  2.63657
-   3 │ Merced       519.25   8.18989e-15
-   4 │ Los Angeles  647.267  47.7685
-   5 │ Orange       710.563  2.21461e-13
+   1 │ Santa Clara  732.077  58.2169
+   2 │ San Diego    659.436  2.66703
+   3 │ Merced       519.25   2.28936e-15
+   4 │ Los Angeles  647.267  47.6233
+   5 │ Orange       710.563  2.19826e-13
    6 │ Fresno       472.0    1.13687e-13
-   7 │ Plumas       709.556  1.26823e-13
-   8 │ Alameda      669.0    1.26888e-13
-   9 │ San Joaquin  551.189  2.17297e-13
+   7 │ Plumas       709.556  1.26058e-13
+   8 │ Alameda      669.0    1.27527e-13
+   9 │ San Joaquin  551.189  2.1791e-13
   10 │ Kern         452.5    0.0
-  11 │ Mendocino    623.25   1.09409e-13
+  11 │ Mendocino    623.25   1.09545e-13
 ```
 """
 function mean(x::Symbol, domain::Symbol, design::ReplicateDesign)

diff --git a/src/ratio.jl b/src/ratio.jl
@@ -8,12 +8,13 @@ julia> apiclus1 = load_data("apiclus1");
 
 julia> clus_one_stage = SurveyDesign(apiclus1; clusters = :dnum, weights = :pw) |> bootweights;
 
-ratio(:api00, :enroll, clus_one_stage)
+julia> ratio(:api00, :enroll, clus_one_stage)
 1×2 DataFrame
  Row │ ratio    SE
      │ Float64  Float64
 ─────┼───────────────────
-   1 │ 1.17182  0.133361
+   1 │ 1.17182  0.131518
+
 ```
 """
 function ratio(variable_num::Symbol, variable_den::Symbol, design::ReplicateDesign)

diff --git a/src/total.jl b/src/total.jl
@@ -13,15 +13,15 @@ julia> total(:api00, clus_one_stage)
  Row │ total      SE
      │ Float64    Float64
 ─────┼──────────────────────
-   1 │ 3.98999e6  9.10443e5
+   1 │ 3.98999e6  9.01611e5
 
 julia> total([:api00, :enroll], clus_one_stage)
 2×3 DataFrame
  Row │ names   total      SE
      │ String  Float64    Float64
 ─────┼──────────────────────────────
-   1 │ api00   3.98999e6  9.10443e5
-   2 │ enroll  3.40494e6  9.47987e5
+   1 │ api00   3.98999e6  9.01611e5
+   2 │ enroll  3.40494e6  9.33396e5
 ```
 """
 function total(x::Symbol, design::ReplicateDesign)
@@ -52,17 +52,17 @@ julia> total(:api00, :cname, clus_one_stage)
  Row │ cname        total           SE
      │ String15     Float64         Any
 ─────┼────────────────────────────────────────
-   1 │ Santa Clara       6.44244e5  4.29558e5
-   2 │ San Diego         1.2276e6   8.60246e5
-   3 │ Merced        70300.2        70757.4
-   4 │ Los Angeles       3.2862e5   2.95688e5
-   5 │ Orange            3.84807e5  3.77128e5
-   6 │ Fresno        63903.1        64455.2
-   7 │ Plumas            2.16147e5  2.12279e5
-   8 │ Alameda      249080.0        2.5221e5
-   9 │ San Joaquin       6.90276e5  6.92353e5
-  10 │ Kern          30631.5        30333.5
-  11 │ Mendocino     84380.6        80774.4
+   1 │ Santa Clara       6.44244e5  4.2273e5
+   2 │ San Diego         1.2276e6   8.62727e5
+   3 │ Merced        70300.2        71336.3
+   4 │ Los Angeles       3.2862e5   2.93936e5
+   5 │ Orange            3.84807e5  3.88014e5
+   6 │ Fresno        63903.1        64781.7
+   7 │ Plumas            2.16147e5  2.12089e5
+   8 │ Alameda      249080.0        2.49228e5
+   9 │ San Joaquin       6.90276e5  6.81604e5
+  10 │ Kern          30631.5        30870.3
+  11 │ Mendocino     84380.6        80215.9
 ```
 """
 function total(x::Symbol, domain::Symbol, design::ReplicateDesign)