preserve county and municipality boundaries as much as possible (1)
Interpretation of requirements
We enforce a maximum population deviation of 0.5%.
We apply the standard algorithmic county constraint.
Data Sources
Data for Maine comes from the Voting and Election Science Team for 2016, 2018, and 2020. It is retabulated to 2020 Census tracts, as 2020 Census VTDs do not cover the majority of Maine's geography.
Pre-processing Notes
Islands tracts were connected to the nearest tract within the same district.
Simulation Notes
We sample 5,000 districting plans for Maine, across 4 independent runs of the SMC algorithm.
We use the standard county constraint.
We weaken the compactness parameter to 0.9 due to the relatively small state size and total number of tracts to encourage more diversity in the sample.
Validation
SMC: 5,000 sampled plans of 2 districts on 401 units
`adapt_k_thresh`=0.985 • `seq_alpha`=0.5
`est_label_mult`=1 • `pop_temper`=0
Plan diversity 80% range: 0.016 to 0.483
✖ WARNING: Low plan diversity
R-hat values for summary statistics:
pop_overlap total_vap plan_dev comp_edge comp_polsby pop_white pop_black pop_hisp pop_aian
1.0013128 1.0011930 1.0012554 1.0001128 1.0009325 1.0004023 1.0000092 0.9999668 1.0021643
pop_asian pop_nhpi pop_other pop_two vap_white vap_black vap_hisp vap_aian vap_asian
1.0017337 1.0025655 1.0022458 1.0000240 1.0004747 1.0000503 1.0008113 1.0023318 1.0019800
vap_nhpi vap_other vap_two pre_16_dem_cli pre_16_rep_tru gov_18_dem_mil gov_18_rep_moo uss_18_dem_rin uss_18_rep_bra
1.0021021 1.0016257 0.9999838 1.0008052 1.0008569 1.0007473 1.0011623 1.0009893 1.0016979
pre_20_dem_bid pre_20_rep_tru uss_20_dem_gid uss_20_rep_col arv_16 adv_16 arv_18 adv_18 arv_20
1.0009598 1.0003298 1.0008160 1.0005918 1.0008569 1.0008052 1.0014489 1.0004953 1.0004344
adv_20 county_splits muni_splits ndv nrv ndshare e_dvs pr_dem e_dem
1.0008716 1.0011363 1.0002822 1.0008561 1.0005056 1.0004675 1.0004234 1.0013797 1.0001447
egap
1.0018267
Sampling diagnostics for SMC run 1 of 4 (1,250 samples)
Eff. samples (%) Acc. rate Log wgt. sd Max. unique Est. k
Split 1 1,096 (87.7%) 9.9% 0.56 779 ( 99%) 5
Resample 791 (63.2%) NA% 0.56 715 ( 90%) NA
Sampling diagnostics for SMC run 2 of 4 (1,250 samples)
Eff. samples (%) Acc. rate Log wgt. sd Max. unique Est. k
Split 1 1,113 (89.1%) 10.3% 0.52 785 ( 99%) 5
Resample 846 (67.7%) NA% 0.52 723 ( 92%) NA
Sampling diagnostics for SMC run 3 of 4 (1,250 samples)
Eff. samples (%) Acc. rate Log wgt. sd Max. unique Est. k
Split 1 1,080 (86.4%) 10.3% 0.56 780 ( 99%) 5
Resample 776 (62.1%) NA% 0.56 705 ( 89%) NA
Sampling diagnostics for SMC run 4 of 4 (1,250 samples)
Eff. samples (%) Acc. rate Log wgt. sd Max. unique Est. k
Split 1 996 (79.7%) 10.0% 0.57 805 (102%) 5
Resample 773 (61.8%) NA% 0.57 699 ( 88%) NA
• Watch out for low effective samples, very low acceptance rates (less than 1%), large std. devs. of the log weights (more than 3 or so),
and low numbers of unique plans. R-hat values for summary statistics should be between 1 and 1.05.
• Low diversity: Check for potential bottlenecks. Increase the number of samples. Examine the diversity plot with
`hist(plans_diversity(plans), breaks=24)`. Consider weakening or removing constraints, or increasing the population tolerance. If the
accpetance rate drops quickly in the final splits, try increasing `pop_temper` by 0.01.
Redistricting requirements
In Maine, following Title 21-A, Chapter 15, Section 1206, districts must:
Interpretation of requirements
We enforce a maximum population deviation of 0.5%. We apply the standard algorithmic county constraint.
Data Sources
Data for Maine comes from the Voting and Election Science Team for 2016, 2018, and 2020. It is retabulated to 2020 Census tracts, as 2020 Census VTDs do not cover the majority of Maine's geography.
Pre-processing Notes
Islands tracts were connected to the nearest tract within the same district.
Simulation Notes
We sample 5,000 districting plans for Maine, across 4 independent runs of the SMC algorithm. We use the standard county constraint. We weaken the compactness parameter to 0.9 due to the relatively small state size and total number of tracts to encourage more diversity in the sample.
Validation
Checklist
TODO
lines from the template code have been removedenforce_style()
to format my coderedist_map
andredist_plans
objects, and summary statistics) have been edited@CoryMcCartan
Note: 2 district state, so I ran 4 runs of 1250 for diversity.