May edition

technocrat · Jun 19, 2024 · 9ae142a · 9ae142a
1 parent f3f813c
commit 9ae142a
Show file tree

Hide file tree

Showing 82 changed files with 732 additions and 201 deletions.
diff --git a/_assets/img/models/AZ_2020.png b/_assets/img/models/AZ_2020.png
diff --git a/_assets/img/models/AZ_apr.png b/_assets/img/models/AZ_apr.png
diff --git a/_assets/img/models/AZ_mar.png b/_assets/img/models/AZ_mar.png
diff --git a/_assets/img/models/AZ_may.png b/_assets/img/models/AZ_may.png
diff --git a/_assets/img/models/GA_2020.png b/_assets/img/models/GA_2020.png
diff --git a/_assets/img/models/GA_apr.png b/_assets/img/models/GA_apr.png
diff --git a/_assets/img/models/GA_mar.png b/_assets/img/models/GA_mar.png
diff --git a/_assets/img/models/MI_2020.png b/_assets/img/models/MI_2020.png
diff --git a/_assets/img/models/MI_apr.png b/_assets/img/models/MI_apr.png
diff --git a/_assets/img/models/MI_mar.png b/_assets/img/models/MI_mar.png
diff --git a/_assets/img/models/MI_may.png b/_assets/img/models/MI_may.png
diff --git a/_assets/img/models/NC_2020.png b/_assets/img/models/NC_2020.png
diff --git a/_assets/img/models/NC_apr.png b/_assets/img/models/NC_apr.png
diff --git a/_assets/img/models/NC_mar.png b/_assets/img/models/NC_mar.png
diff --git a/_assets/img/models/NC_may.png b/_assets/img/models/NC_may.png
diff --git a/_assets/img/models/NV_2020.png b/_assets/img/models/NV_2020.png
diff --git a/_assets/img/models/NV_mar.png b/_assets/img/models/NV_mar.png
diff --git a/_assets/img/models/NV_may.png b/_assets/img/models/NV_may.png
diff --git a/_assets/img/models/PA_2020.png b/_assets/img/models/PA_2020.png
diff --git a/_assets/img/models/PA_apr.png b/_assets/img/models/PA_apr.png
diff --git a/_assets/img/models/PA_mar.png b/_assets/img/models/PA_mar.png
diff --git a/_assets/img/models/PA_may.png b/_assets/img/models/PA_may.png
diff --git a/_assets/img/models/WI_2020.png b/_assets/img/models/WI_2020.png
diff --git a/_assets/img/models/WI_apr.png b/_assets/img/models/WI_apr.png
diff --git a/_assets/img/models/WI_mar.png b/_assets/img/models/WI_mar.png
diff --git a/_assets/img/models/WI_may.png b/_assets/img/models/WI_may.png
diff --git a/_assets/meta_cohort_2020.bson b/_assets/meta_cohort_2020.bson
diff --git a/_assets/meta_cohort_2022.bson b/_assets/meta_cohort_2022.bson
diff --git a/_assets/objs/AZ_2020_p_sample.bson b/_assets/objs/AZ_2020_p_sample.bson
diff --git a/_assets/objs/AZ_apr_p_sample.bson b/_assets/objs/AZ_apr_p_sample.bson
diff --git a/_assets/objs/AZ_mar_p_sample.bson b/_assets/objs/AZ_mar_p_sample.bson
diff --git a/_assets/objs/AZ_may_p_sample.bson b/_assets/objs/AZ_may_p_sample.bson
diff --git a/_assets/objs/GA_2020_p_sample.bson b/_assets/objs/GA_2020_p_sample.bson
diff --git a/_assets/objs/GA_apr_p_sample.bson b/_assets/objs/GA_apr_p_sample.bson
diff --git a/_assets/objs/GA_mar_p_sample.bson b/_assets/objs/GA_mar_p_sample.bson
diff --git a/_assets/objs/MI_2020_p_sample.bson b/_assets/objs/MI_2020_p_sample.bson
diff --git a/_assets/objs/MI_apr_p_sample.bson b/_assets/objs/MI_apr_p_sample.bson
diff --git a/_assets/objs/MI_mar_p_sample.bson b/_assets/objs/MI_mar_p_sample.bson
diff --git a/_assets/objs/MI_may_p_sample.bson b/_assets/objs/MI_may_p_sample.bson
diff --git a/_assets/objs/NC_2020_p_sample.bson b/_assets/objs/NC_2020_p_sample.bson
diff --git a/_assets/objs/NC_apr_p_sample.bson b/_assets/objs/NC_apr_p_sample.bson
diff --git a/_assets/objs/NC_mar_p_sample.bson b/_assets/objs/NC_mar_p_sample.bson
diff --git a/_assets/objs/NC_may_p_sample.bson b/_assets/objs/NC_may_p_sample.bson
diff --git a/_assets/objs/NV_2020_p_sample.bson b/_assets/objs/NV_2020_p_sample.bson
diff --git a/_assets/objs/NV_mar_p_sample.bson b/_assets/objs/NV_mar_p_sample.bson
diff --git a/_assets/objs/NV_may_p_sample.bson b/_assets/objs/NV_may_p_sample.bson
diff --git a/_assets/objs/PA_2020_p_sample.bson b/_assets/objs/PA_2020_p_sample.bson
diff --git a/_assets/objs/PA_apr_p_sample.bson b/_assets/objs/PA_apr_p_sample.bson
diff --git a/_assets/objs/PA_mar_p_sample.bson b/_assets/objs/PA_mar_p_sample.bson
diff --git a/_assets/objs/PA_may_p_sample.bson b/_assets/objs/PA_may_p_sample.bson
diff --git a/_assets/objs/WI_2020_p_sample.bson b/_assets/objs/WI_2020_p_sample.bson
diff --git a/_assets/objs/WI_apr_p_sample.bson b/_assets/objs/WI_apr_p_sample.bson
diff --git a/_assets/objs/WI_mar_p_sample.bson b/_assets/objs/WI_mar_p_sample.bson
diff --git a/_assets/objs/WI_may_p_sample.bson b/_assets/objs/WI_may_p_sample.bson
diff --git a/_assets/election_priors.csv → _assets/objs/election_priors b/_assets/election_priors.csv → _assets/objs/election_priors
@@ -1,4 +1,4 @@
-st,biden_pop,tot
+st,num_wins,num_votes
 AZ,1672143,3333829
 GA,2473633,4935487
 MI,2804040,5453892

diff --git a/_assets/objs/may_polls.bson b/_assets/objs/may_polls.bson
diff --git a/_assets/scripts/election_priors.jl b/_assets/scripts/election_priors.jl
@@ -8,4 +8,4 @@ election_priors        = election_priors[:,[1,2,3]]
 election_priors.tot    = election_priors.biden_pop .+ election_priors.trump_pop
 election_priors        = election_priors[:,[:st,:biden_pop,:tot]]
 rename!(election_priors,["st","num_wins","num_votes"])
-# CSV.write("../objs/election_priors.jl",election_priors)
+# CSV.write("../objs/election_priors",election_priors)
diff --git a/_assets/scripts/first_posterior.jl b/_assets/scripts/first_posterior.jl
@@ -2,43 +2,43 @@ include("first_posterior_forepart.jl")
 # Set the number of votes and wins
 last_election = CSV.read("../objs/election_priors.csv", DataFrame)
 
-
-ST        = last_election[1, :st]
-num_wins  = last_election[1, :num_wins]
-num_votes = last_election[1, :num_votes]
+#------------------------------------------------------------------
+# row indexed to state in alpha order; repeat for each of 1:7
+#------------------------------------------------------------------
+ST        = last_election[4, :st]
+num_wins  = last_election[4, :num_wins]
+num_votes = last_election[4, :num_votes]
 margins   = CSV.read("../objs/margins.csv", DataFrame)
 margin    = first(margins[margins.st .== ST, :pct])
 
-chain  = sample(election_model(num_votes, num_wins), sampler, num_samples, init_params=init_params)
-
-p_intv = quantile(chain[:p], [0.025, 0.975])
-p_mean = summarystats(chain)[1, :mean]
-p_mcse = summarystats(chain)[1, :mcse]
-p_rhat = summarystats(chain)[1, :rhat]
-p_df   = DataFrame(median = median(chain[:p]),
-                   mean   = mean(chain[:p]),
-                   mode   = mode(chain[:p]),
-                   q025   = p_intv[1],
-                   q975   = p_intv[2],
-                   mcse   = summarystats(chain)[1, :mcse],
-                   rhat   = summarystats(chain)[1, :rhat])
-
-# Extract the :p parameter from the chain object
-p_samples = chain[:p]
-
-# Flatten the p_samples array into a 1D vector
-p_vec = vec(p_samples)
-
-# Compute the density estimate
+chain     = sample(election_model(num_votes, num_wins), sampler, 
+                   num_samples, init_params=init_params)
+
+p_intv    = quantile(chain[:p], [0.025, 0.975])
+p_mean    = summarystats(chain)[1, :mean]
+p_mcse    = summarystats(chain)[1, :mcse]
+p_rhat    = summarystats(chain)[1, :rhat]
+p_df      = DataFrame(median = median(chain[:p]),
+                      mean   = mean(chain[:p]),
+                      mode   = mode(chain[:p]),
+                      q025   = p_intv[1],
+                      q975   = p_intv[2],
+                      mcse   = summarystats(chain)[1, :mcse],
+                      rhat   = summarystats(chain)[1, :rhat])
+
+p_samples  = chain[:p]
+p_vec      = vec(p_samples)
 kde_result = kde(p_vec)
 
 include("first_posterior_aftpart.jl")
-# Display posterior density plot
+
 posterior_interval = p_intv
-fig = draw_density()
-deep = deepcopy(chain)
+fig                = draw_density()
+deep               = deepcopy(chain)
+
 @save ("../objs/$ST" * "_2020_p_sample.bson") deep
 save(("../img/models/$ST" * "_2020.png"), fig)
+
 out       = Vector(p_df[1,:])
 out       = round.(out,digits = 4)
 p_df[1,:] = out

diff --git a/_assets/scripts/first_posterior_aftpart.jl b/_assets/scripts/first_posterior_aftpart.jl
@@ -31,7 +31,7 @@ function draw_density()
     fig = Figure(size = (600, 400))
 
     # Add an axis to the figure
-    ax = Axis(fig[1, 1], xlabel = "Likelihood of Biden win", ylabel = "Number of draws", title = "Model: Biden wins in $ST from 2020 election before polling")
+    ax = Axis(fig[1, 1], xlabel = "Likelihood of Biden win", ylabel = "Number of draws", title = "Model: Biden results in $ST from 2020 election before polling")
 
     # Plot the full density curve
     lines!(ax, kde_result.x, kde_result.density, color = "#a3b35c", linewidth = 3, strokewidth = 4, strokecolor = GREENBAR, label = "Draws")

diff --git a/_assets/scripts/first_posterior_forepart.jl b/_assets/scripts/first_posterior_forepart.jl
@@ -23,20 +23,16 @@ const LORANGE  = "rgb(225, 170, 110)"
 # Define the model
 @model function election_model(num_votes::Int64, num_wins::Int64)
     # Prior: Beta(2, 2) equivalent to a close race going in
-    p ~ Beta(2, 2)
-
-    # Likelihood: Binomial(num_votes, p)
+    p ~ Beta(2,2)
+    # Likelihood
     num_wins ~ Binomial(num_votes, p)
 end
 
-# Set up the sampler
 sampler = NUTS(0.65)
-
-# Specify the number of samples and chains
 num_samples = 10000
 num_chains  = 4
 
-# Sample from the posterior
+
 init_params = [Dict(:p => 0.5)]
 
 Month_names = Dict(

diff --git a/_assets/scripts/insert_may.jl b/_assets/scripts/insert_may.jl
@@ -7,7 +7,7 @@ using Statistics
 using StatsPlots
 using Turing
 
-const DESIGN_ERROR = 0.012
+# const DESIGN_ERROR = 0.012
 
 @enum Month mar apr may jun jul aug sep oct nov
 
@@ -28,14 +28,18 @@ const DESIGN_ERROR = 0.012
     fo3
     hi2
     hi3
+    hp2
     ma2
     ma3
     mi2
     mi3
     mr2
     mr3
+    ny2
     qi2
     qi3
+    si2
+    si3
     sp2
     sp3
     su2
@@ -58,28 +62,30 @@ end
 
 @load "../objs/apr_polls.bson" months
 
-"""
-months[may][AZ][bl2] = [Poll(43,49, 801)]
-months[may][AZ][ec2] = [Poll(44,48,1000)]
-months[may][GA][bl2] = [Poll(45,47, 802)]
-months[may][GA][ec2] = [Poll(44,47,1000)]
-months[may][GA][fo2] = [Poll(45,51,1128)]
-months[may][MI][bi2] = [Poll(43,51, 708)]
-months[may][MI][ec2] = [Poll(44,45,1000)]
-months[may][MI][fo2] = [Poll(46,49,1106)]
-months[may][MI][mr2] = [Poll(46,42, 600)]
-months[may][NC][ec2] = [Poll(42,47,1000)]
-months[may][NC][bl2] = [Poll(41,51, 703)]
-months[may][NC][ma2] = [Poll(43,49, 645)]
-months[may][NC][qi2] = [Poll(46,48,1401)]
-months[may][NV][ec2] = [Poll(44,45,1000)]
-months[may][PA][bi2] = [Poll(46,47, 803)]
-months[may][PA][ec2] = [Poll(45,47,1000)]
-months[may][WI][bl2] = [Poll(44,48, 703)]
-months[may][WI][ec2] = [Poll(45,47,1000)]
-"""
+months[may][AZ][bl2] = [Poll(44,49, 795)]
+months[may][GA][bl2] = [Poll(44,47, 795)]
+months[may][MI][bl3] = [Poll(40,44, 704)]
+months[may][NC][bl2] = [Poll(42,49, 704)]
+months[may][WI][bl2] = [Poll(46,47, 693)]
+months[may][NV][bl2] = [Poll(47,47, 459)]
+months[may][PA][bl2] = [Poll(46,48, 812)]
+months[may][AZ][ec2] = [Poll(42,49, 626)]
+months[may][GA][ec2] = [Poll(39,49, 604)]
+months[may][MI][ec2] = [Poll(36,38, 616)]
+months[may][NV][ec2] = [Poll(38,50, 614)]
+months[may][PA][ec2] = [Poll(44,47,1023)]
+months[may][WI][ec2] = [Poll(47,45, 614)]
+months[may][NC][hp2] = [Poll(43,48,1002)]
 months[may][WI][qi2] = [Poll(50,44,1497)]
+months[may][AZ][si2] = [Poll(42,49, 626)]
+months[may][GA][si2] = [Poll(39,49, 604)]
+months[may][MI][si3] = [Poll(36,38, 616)]
+months[may][NV][si2] = [Poll(38,50, 614)]
+months[may][PA][si2] = [Poll(44,47,1023)]
+months[may][WI][si2] = [Poll(47,45, 614)]
 
-# @save "../objs/may_polls.bson" months
+
+# needs to be done manually rather than with include()
+#@save "../objs/may_polls.bson" months
 
 
diff --git a/_assets/scripts/mar_poll.jl b/_assets/scripts/mar_poll.jl
@@ -0,0 +1,20 @@
+@enum State PA GA NC MI AZ WI NV
+@enum Month mar apr may jun jul aug sep oct
+
+
+STATE       = State
+prior_month = "mar"
+mon         = mar
+MON         = "mar"
+Mon         = "mar"
+st          = "NV"
+ST          =  NV
+
+include("mar_poll_head.jl")
+prior_poll = BSON.load("../objs/"*"$st"*"_2020_p_sample.bson")
+@load "../objs/"*"$MON"*"_Polls.bson" months
+
+current_month = remove_empties(months[mon])
+
+include("polls_foot.jl")
+
diff --git a/_assets/scripts/mar_poll_foot.jl b/_assets/scripts/mar_poll_foot.jl
@@ -0,0 +1,72 @@
+
+@model function poll_model(num_votes::Int64, num_wins::Int64, prior_dist::Distribution)
+    # Define the prior using the informed prior distribution
+    p ~ prior_dist
+    # Define the likelihood with additional uncertainty
+    num_wins ~ Binomial(num_votes, p)
+end
+
+processed_polls        = Dict(state    => 
+                         Dict(pollster => 
+                         process_polls(polls) for (pollster, polls) in pollsters) 
+                         for (state, pollsters) in current_month)
+
+processed_polls_totals = Dict(state      => 
+                         Dict("num_wins" => 
+                         sum(first(values(polls)) for polls in values(pollsters)),
+                         "num_votes"     => 
+                         sum(last(values(polls)) for polls in values(pollsters)))
+                         for (state, pollsters) in processed_polls)
+
+num_wins               = processed_polls_totals[ST]["num_wins"]
+num_votes              = processed_polls_totals[ST]["num_votes"]
+
+poll_posterior         = prior_poll
+
+posterior_mean         = mean(poll_posterior[:deep][:p])
+posterior_var          =  var(poll_posterior[:deep][:p])
+prior_alpha            = posterior_mean * 
+                        (posterior_mean * (1 - posterior_mean) / posterior_var - 1)
+prior_beta             = (1 - posterior_mean) * (posterior_mean *
+                         (1 - posterior_mean) / posterior_var - 1)
+prior_dist             = Beta(prior_alpha, prior_beta)
+
+model                  = poll_model(num_votes, num_wins, prior_dist)
+sampler                = NUTS(0.65)
+num_samples            = 10000
+num_chains             = 4
+init_params            = [Dict(:p => 0.5)]
+chain                  = sample(poll_model(num_votes, num_wins, prior_dist), 
+                         sampler, num_samples, init_params=init_params)
+
+p_intv = quantile(chain[:p], [0.025, 0.975])
+p_mean = summarystats(chain)[1,:mean]
+p_mcse = summarystats(chain)[1,:mcse]
+p_rhat = summarystats(chain)[1,:rhat]
+p_df   = DataFrame(median = median(chain[:p]),
+                   mean   = mean(chain[:p]),
+                   mode   = mode(chain[:p]),
+                   q025   = p_intv[1],
+                   q975   = p_intv[2],
+                   mcse   = summarystats(chain)[1,:mcse],
+                   rhat   = summarystats(chain)[1,:rhat])
+
+p_samples  = chain[:p]
+p_vec      = vec(p_samples)
+kde_result = kde(p_vec)
+
+posterior_interval = p_intv
+fig                = draw_density()
+save(("../img/models/"*"$st"*"_"*"$mon"*".png"), fig)
+
+deep = deepcopy(chain)
+@save "../objs/"*"$st"*"_"*"$mon"*"_p_sample.bson" deep
+
+out       = Vector(p_df[1,:])
+out       = round.(out,digits = 4)
+p_df[1,:] = out
+pretty_table(p_df,backend=Val(:html),show_subheader = false)
+chain
+summarystats(chain)
+autocor(chain)
+hpd(chain)