rm(list=ls())
library(ergm)
## Loading required package: network
## network: Classes for Relational Data
## Version 1.15 created on 2019-04-01.
## copyright (c) 2005, Carter T. Butts, University of California-Irvine
## Mark S. Handcock, University of California -- Los Angeles
## David R. Hunter, Penn State University
## Martina Morris, University of Washington
## Skye Bender-deMoll, University of Washington
## For citation information, type citation("network").
## Type help("network-package") to get started.
##
## ergm: version 3.10.4, created on 2019-06-10
## Copyright (c) 2019, Mark S. Handcock, University of California -- Los Angeles
## David R. Hunter, Penn State University
## Carter T. Butts, University of California -- Irvine
## Steven M. Goodreau, University of Washington
## Pavel N. Krivitsky, University of Wollongong
## Martina Morris, University of Washington
## with contributions from
## Li Wang
## Kirk Li, University of Washington
## Skye Bender-deMoll, University of Washington
## Chad Klumb
## Based on "statnet" project software (statnet.org).
## For license and citation information see statnet.org/attribution
## or type citation("ergm").
## NOTE: Versions before 3.6.1 had a bug in the implementation of the
## bd() constriant which distorted the sampled distribution somewhat.
## In addition, Sampson's Monks datasets had mislabeled vertices. See
## the NEWS and the documentation for more details.
## NOTE: Some common term arguments pertaining to vertex attribute
## and level selection have changed in 3.10.0. See terms help for
## more details. Use 'options(ergm.term=list(version="3.9.4"))' to
## use old behavior.
library(network)
library(networkDynamic)
##
## networkDynamic: version 0.10.0, created on 2019-04-04
## Copyright (c) 2019, Carter T. Butts, University of California -- Irvine
## Ayn Leslie-Cook, University of Washington
## Pavel N. Krivitsky, University of Wollongong
## Skye Bender-deMoll, University of Washington
## with contributions from
## Zack Almquist, University of California -- Irvine
## David R. Hunter, Penn State University
## Li Wang
## Kirk Li, University of Washington
## Steven M. Goodreau, University of Washington
## Jeffrey Horner
## Martina Morris, University of Washington
## Based on "statnet" project software (statnet.org).
## For license and citation information see statnet.org/attribution
## or type citation("networkDynamic").
library(tergm)
##
## tergm: version 3.6.1, created on 2019-06-12
## Copyright (c) 2019, Pavel N. Krivitsky, University of Wollongong
## Mark S. Handcock, University of California -- Los Angeles
## with contributions from
## David R. Hunter, Penn State University
## Steven M. Goodreau, University of Washington
## Martina Morris, University of Washington
## Nicole Bohme Carnegie, New York University
## Carter T. Butts, University of California -- Irvine
## Ayn Leslie-Cook, University of Washington
## Skye Bender-deMoll
## Li Wang
## Kirk Li, University of Washington
## Based on "statnet" project software (statnet.org).
## For license and citation information see statnet.org/attribution
## or type citation("tergm").
load("init-pop-atts.RData")
formation <- ~edges+degree(0:1)+
nodemix("identity", levels=c("msm", "tgw", "hrh"))+
nodemix("age_group", levels = c("18-24", "25-34", "35-44", "45-64"))+
nodemix("race", levels=c("White", "Black", "Hispanic"))
dissolution <- ~offset(edges)
constraints <- ~.
formation.casual.net <- update.formula(formation, casual.net~.)
casual.net <- net
casual_mean_degree <- 1.33
casual_mean_nedges <- n*casual_mean_degree/2
casual.degprop.0 <- 0.167763158
casual.degprop.1 <- 0.430921053
casual.degprop.2 <- 0.401315789
casual.deg.prop <- c(casual.degprop.0, casual.degprop.1, casual.degprop.2)
casual.deg.nodes <- n*casual.deg.prop
casual.pt.duration <- 313.8824 #by estimation
min.age <- 18; max.age <- 64 #by design
d.rate <- 1/((max.age - min.age)*365) #derived
pg.casual <- (casual.pt.duration - 1)/casual.pt.duration #derived
ps2 <- (1 - d.rate)^2 #derived
theta.casual.diss <- log(pg.casual/(ps2-pg.casual)) #derived
casual.msm.msm.prop <- 0.835361469
casual.tgw.tgw.prop <- 0.005998365
casual.msm.tgw.prop <- (0.054821125+0.045494773)/2
casual.msm.hrh.prop <- 0
casual.tgw.hrh.prop <- 0.055500534
casual.hrh.hrh.prop <- 0
casual.18.24.18.24 <- 0.060648808
casual.25.34.25.34 <- 0.196050785
casual.35.44.35.44 <- 0.093088858
casual.45.64.45.64 <- 0.23695346
casual.25.34.18.24 <- (0.084626222+0.029619181)/2
casual.35.44.18.24 <- 0
casual.45.64.18.24 <- 0
casual.35.44.25.34 <- (0.080394922+0.062059227)/2
casual.45.64.25.34 <- 0
casual.45.64.35.44 <- (0.050775745+0.088857555)/2
casual.white.white <- 0.162895937
casual.black.white <- (0.095022614 + 0.05279034)/2
casual.hispanic.white <- (0.122171947 + 0.105580681)/2
casual.black.black <- 0.257918548
casual.hispanic.black <- mean(c(0.01809956, 0.036199101))
casual.hispanic.hispanic <- 0.149321272
casual.nodemix.identity <- casual_mean_nedges*(c(casual.msm.msm.prop,
casual.msm.tgw.prop, casual.tgw.tgw.prop,
casual.msm.hrh.prop, casual.tgw.hrh.prop, casual.hrh.hrh.prop
))
casual.nodemix.age <- casual_mean_nedges*c(
casual.18.24.18.24,
casual.25.34.18.24, casual.25.34.25.34,
casual.35.44.18.24, casual.35.44.25.34, casual.35.44.35.44,
casual.45.64.18.24, casual.45.64.25.34, casual.45.64.35.44, casual.45.64.45.64
)
casual.nodemix.race <- casual_mean_nedges*c(casual.white.white,
casual.black.white, casual.black.black,
casual.hispanic.white, casual.hispanic.black, casual.hispanic.hispanic)
target.stats <- c(casual_mean_nedges, casual.deg.nodes[1:2], casual.nodemix.identity,
casual.nodemix.age, casual.nodemix.race)
casual.fit <- ergm(
formation.casual.net,
target.stats=target.stats,
constraints=constraints,
eval.loglik=FALSE,
verbose=TRUE,
control=control.ergm(MCMLE.maxit=500)
)
## Evaluating network in model.
## Initializing Metropolis-Hastings proposal(s): ergm:MH_TNT
## Initializing model.
## Warning: `set_attrs()` is deprecated as of rlang 0.3.0
## This warning is displayed once per session.
## Using initial method 'MPLE'.
## Constructing an approximate response network.
## Starting 4 MCMC iterations of 131072 steps each.
## #1 of 4:
## SAN Metropolis-Hastings accepted 48.193% of 8192 proposed steps.
## SAN summary statistics:
## edges degree0
## 3195 1113
## degree1 mix.identity.msm.msm
## 2199 0
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 0 0
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0 0
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0 101
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 278 331
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 218 351
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 172 324
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 481 428
## mix.age_group.45-64.45-64 mix.race.White.White
## 511 565
## mix.race.White.Black mix.race.Black.Black
## 632 426
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 693 509
## mix.race.Hispanic.Hispanic
## 370
## Meanstats Goal:
## edges degree0
## 3325.00000 838.81579
## degree1 mix.identity.msm.msm
## 2154.60527 2777.57688
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 166.77518 19.94456
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0.00000 184.53928
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0.00000 201.65729
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 189.93298 651.86886
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 0.00000 236.83002
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 309.52045 0.00000
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 0.00000 232.14036
## mix.age_group.45-64.45-64 mix.race.White.White
## 787.87025 541.62899
## mix.race.White.Black mix.race.Black.Black
## 245.73904 857.57917
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 378.63874 90.27152
## mix.race.Hispanic.Hispanic
## 496.49323
## Difference: SAN target.stats - Goal target.stats =
## [1] -130.00000 274.18421 44.39473 -2777.57688 -166.77518
## [6] -19.94456 0.00000 -184.53928 0.00000 -100.65729
## [11] 88.06702 -320.86886 218.00000 114.16998 -137.52045
## [16] 324.00000 481.00000 195.85964 -276.87025 23.37101
## [21] 386.26096 -431.57917 314.36126 418.72848 -126.49323
## New statistics scaling =
## [1] 0.02869715 0.06040529 0.02007849 0.01957396 0.01957396 0.01957396
## [7] 0.01957396 0.01957396 0.01957396 0.15884464 0.05511463 0.04857241
## [13] 0.05767377 0.04504002 0.10257355 0.03563637 0.02622357 0.03402426
## [19] 0.03496066 0.02832223 0.01957396 0.03919900 0.02159292 0.02655125
## [25] 0.03947204
## Scaled Mahalanobis distance = 211598.315923159
## #2 of 4:
## SAN Metropolis-Hastings accepted 27.700% of 17408 proposed steps.
## SAN summary statistics:
## edges degree0
## 3339 795
## degree1 mix.identity.msm.msm
## 2097 0
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 0 0
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0 0
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0 247
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 246 688
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 91 305
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 366 132
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 137 305
## mix.age_group.45-64.45-64 mix.race.White.White
## 822 649
## mix.race.White.Black mix.race.Black.Black
## 405 909
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 533 250
## mix.race.Hispanic.Hispanic
## 593
## Meanstats Goal:
## edges degree0
## 3325.00000 838.81579
## degree1 mix.identity.msm.msm
## 2154.60527 2777.57688
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 166.77518 19.94456
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0.00000 184.53928
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0.00000 201.65729
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 189.93298 651.86886
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 0.00000 236.83002
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 309.52045 0.00000
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 0.00000 232.14036
## mix.age_group.45-64.45-64 mix.race.White.White
## 787.87025 541.62899
## mix.race.White.Black mix.race.Black.Black
## 245.73904 857.57917
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 378.63874 90.27152
## mix.race.Hispanic.Hispanic
## 496.49323
## Difference: SAN target.stats - Goal target.stats =
## [1] 14.00000 -43.81579 -57.60527 -2777.57688 -166.77518
## [6] -19.94456 0.00000 -184.53928 0.00000 45.34271
## [11] 56.06702 36.13114 91.00000 68.16998 56.47955
## [16] 132.00000 137.00000 72.85964 34.12975 107.37101
## [21] 159.26096 51.42083 154.36126 159.72848 96.50677
## New statistics scaling =
## [1] 0.01887423 0.05110824 0.01395073 0.01395073 0.01395073 0.01395073
## [7] 0.01395073 0.01395073 0.01395073 0.12390522 0.06666908 0.04507374
## [13] 0.09433830 0.05040252 0.07294739 0.05989546 0.04414356 0.04448050
## [19] 0.03976380 0.03131724 0.02720901 0.03217704 0.02758583 0.03316524
## [25] 0.03928851
## Scaled Mahalanobis distance = 109703.29317582
## #3 of 4:
## SAN Metropolis-Hastings accepted 21.892% of 34816 proposed steps.
## SAN summary statistics:
## edges degree0
## 3177 841
## degree1 mix.identity.msm.msm
## 2160 0
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 0 0
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0 0
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0 256
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 247 706
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 57 293
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 365 58
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 60 290
## mix.age_group.45-64.45-64 mix.race.White.White
## 845 634
## mix.race.White.Black mix.race.Black.Black
## 343 949
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 475 187
## mix.race.Hispanic.Hispanic
## 589
## Meanstats Goal:
## edges degree0
## 3325.00000 838.81579
## degree1 mix.identity.msm.msm
## 2154.60527 2777.57688
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 166.77518 19.94456
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0.00000 184.53928
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0.00000 201.65729
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 189.93298 651.86886
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 0.00000 236.83002
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 309.52045 0.00000
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 0.00000 232.14036
## mix.age_group.45-64.45-64 mix.race.White.White
## 787.87025 541.62899
## mix.race.White.Black mix.race.Black.Black
## 245.73904 857.57917
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 378.63874 90.27152
## mix.race.Hispanic.Hispanic
## 496.49323
## Difference: SAN target.stats - Goal target.stats =
## [1] -148.000000 2.184210 5.394735 -2777.576884 -166.775180
## [6] -19.944564 0.000000 -184.539276 0.000000 54.342713
## [11] 57.067018 54.131140 57.000000 56.169977 55.479547
## [16] 58.000000 60.000000 57.859639 57.129746 92.371009
## [21] 97.260964 91.420828 96.361256 96.728476 92.506771
## New statistics scaling =
## [1] 0.02699081 0.06842869 0.01751413 0.01751413 0.01751413 0.01751413
## [7] 0.01751413 0.01751413 0.01751413 0.09025133 0.06055240 0.03847262
## [13] 0.09973926 0.04714080 0.06818121 0.07215723 0.04841551 0.04063017
## [19] 0.03060359 0.03115669 0.02944298 0.02978256 0.02706391 0.03493857
## [25] 0.03345279
## Scaled Mahalanobis distance = 136217.187690217
## #4 of 4:
## SAN Metropolis-Hastings accepted 0.056% of 69632 proposed steps.
## SAN summary statistics:
## edges degree0
## 3174 839
## degree1 mix.identity.msm.msm
## 2155 0
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 0 0
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0 0
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0 258
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 246 708
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 56 293
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 366 57
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 57 289
## mix.age_group.45-64.45-64 mix.race.White.White
## 844 635
## mix.race.White.Black mix.race.Black.Black
## 340 952
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 473 184
## mix.race.Hispanic.Hispanic
## 590
## Meanstats Goal:
## edges degree0
## 3325.00000 838.81579
## degree1 mix.identity.msm.msm
## 2154.60527 2777.57688
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 166.77518 19.94456
## mix.identity.msm.hrh mix.identity.tgw.hrh
## 0.00000 184.53928
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## 0.00000 201.65729
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 189.93298 651.86886
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## 0.00000 236.83002
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 309.52045 0.00000
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## 0.00000 232.14036
## mix.age_group.45-64.45-64 mix.race.White.White
## 787.87025 541.62899
## mix.race.White.Black mix.race.Black.Black
## 245.73904 857.57917
## mix.race.White.Hispanic mix.race.Black.Hispanic
## 378.63874 90.27152
## mix.race.Hispanic.Hispanic
## 496.49323
## Difference: SAN target.stats - Goal target.stats =
## [1] -151.000000 0.184210 0.394735 -2777.576884 -166.775180
## [6] -19.944564 0.000000 -184.539276 0.000000 56.342713
## [11] 56.067018 56.131140 56.000000 56.169977 56.479547
## [16] 57.000000 57.000000 56.859639 56.129746 93.371009
## [21] 94.260964 94.420828 94.361256 93.728476 93.506771
## New statistics scaling =
## [1] 0.02523285 0.05906940 0.01696120 0.01696120 0.01696120 0.01696120
## [7] 0.01696120 0.01696120 0.01696120 0.09402871 0.05643041 0.03766932
## [13] 0.10649135 0.05083558 0.06099647 0.07561060 0.05173740 0.04600315
## [19] 0.03114238 0.03268255 0.02580873 0.02592061 0.02920382 0.03857442
## [25] 0.03383383
## Scaled Mahalanobis distance = 131910.788011507
## Finished SAN run.
## Observed statistic(s) mix.identity.msm.hrh, mix.identity.hrh.hrh, mix.age_group.18-24.35-44, mix.age_group.18-24.45-64, and mix.age_group.25-34.45-64 are at their smallest attainable values. Their coefficients will be fixed at -Inf.
## Fitting initial model.
## Unable to match target stats. Using MCMLE estimation.
## Starting maximum pseudolikelihood estimation (MPLE):
## Evaluating the predictor and response matrix.
## MPLE covariate matrix has 618 rows.
## Maximizing the pseudolikelihood.
## Finished MPLE.
## Starting Monte Carlo maximum likelihood estimation (MCMLE):
## Density guard set to 63751 from an initial count of 3174 edges.
##
## Iteration 1 of at most 500 with parameter:
## edges degree0
## -10.9524149 -4.8897242
## degree1 mix.identity.msm.msm
## -2.1350850 0.0000000
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 0.0000000 0.0000000
## mix.identity.msm.hrh mix.identity.tgw.hrh
## -Inf 0.0000000
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## -Inf -0.1981090
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## -0.8303186 0.9785100
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## -Inf -0.5435985
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## 0.4040808 -Inf
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## -Inf -1.1033704
## mix.age_group.45-64.45-64 mix.race.White.White
## 0.0000000 -0.5541800
## mix.race.White.Black mix.race.Black.Black
## -0.9856249 1.4634982
## mix.race.White.Hispanic mix.race.Black.Hispanic
## -1.2180623 -1.3143495
## mix.race.Hispanic.Hispanic
## 0.0000000
## Starting unconstrained MCMC...
## Back from unconstrained MCMC.
## Average estimating function values:
## edges degree0
## -198.45801 39.84534
## degree1 mix.identity.msm.msm
## 6.50704 -2777.57688
## mix.identity.msm.tgw mix.identity.tgw.tgw
## -166.77518 -19.94456
## mix.identity.tgw.hrh mix.age_group.18-24.18-24
## -184.53928 78.23139
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 72.99573 62.88505
## mix.age_group.25-34.35-44 mix.age_group.35-44.35-44
## 66.62213 74.26080
## mix.age_group.35-44.45-64 mix.age_group.45-64.45-64
## 79.00417 82.64830
## mix.race.White.White mix.race.White.Black
## 86.25675 106.99827
## mix.race.Black.Black mix.race.White.Hispanic
## 53.43059 89.82805
## mix.race.Black.Hispanic mix.race.Hispanic.Hispanic
## 93.66891 86.00872
## Starting MCMLE Optimization...
## Optimizing with step length 0.00574419914642821.
## Using lognormal metric (see control.ergm function).
## Using log-normal approx (no optim)
## The log-likelihood improved by 12.23.
##
## Iteration 2 of at most 500 with parameter:
## edges degree0
## -4.974382e+00 -4.839849e+00
## degree1 mix.identity.msm.msm
## -2.111492e+00 -8.962179e-13
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 8.290435e-12 5.776338e-12
## mix.identity.msm.hrh mix.identity.tgw.hrh
## -Inf -1.528315e-12
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## -Inf -7.142393e+00
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## -7.831376e+00 -5.975633e+00
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## -Inf -7.483627e+00
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## -6.502442e+00 -Inf
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## -Inf -8.030991e+00
## mix.age_group.45-64.45-64 mix.race.White.White
## -6.920371e+00 4.399702e-01
## mix.race.White.Black mix.race.Black.Black
## -2.361523e-02 2.494891e+00
## mix.race.White.Hispanic mix.race.Black.Hispanic
## -2.043396e-01 -3.353282e-01
## mix.race.Hispanic.Hispanic
## 9.977370e-01
## Starting unconstrained MCMC...
## Back from unconstrained MCMC.
## Average estimating function values:
## edges degree0
## -198.132812 46.652960
## degree1 mix.identity.msm.msm
## -2.150187 -2777.576884
## mix.identity.msm.tgw mix.identity.tgw.tgw
## -166.775180 -19.944564
## mix.identity.tgw.hrh mix.age_group.18-24.18-24
## -184.539276 87.207948
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 51.116822 66.768835
## mix.age_group.25-34.35-44 mix.age_group.35-44.35-44
## 64.225641 81.087946
## mix.age_group.35-44.45-64 mix.age_group.45-64.45-64
## 71.328389 95.311386
## mix.race.White.White mix.race.White.Black
## 88.519447 86.317605
## mix.race.Black.Black mix.race.White.Hispanic
## 75.404226 91.290943
## mix.race.Black.Hispanic mix.race.Hispanic.Hispanic
## 89.525351 85.458919
## Starting MCMLE Optimization...
## Optimizing with step length 0.44024249746706.
## Using lognormal metric (see control.ergm function).
## Using log-normal approx (no optim)
## The log-likelihood improved by 3.057.
##
## Iteration 3 of at most 500 with parameter:
## edges degree0
## -5.176573e+00 -5.178702e+00
## degree1 mix.identity.msm.msm
## -2.255027e+00 -8.964750e-13
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 8.290047e-12 5.776833e-12
## mix.identity.msm.hrh mix.identity.tgw.hrh
## -Inf -1.528748e-12
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## -Inf -7.276748e+00
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## -7.803094e+00 -5.912920e+00
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## -Inf -7.474048e+00
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## -6.537929e+00 -Inf
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## -Inf -8.075933e+00
## mix.age_group.45-64.45-64 mix.race.White.White
## -7.008353e+00 3.645399e-01
## mix.race.White.Black mix.race.Black.Black
## -1.982514e-02 2.537183e+00
## mix.race.White.Hispanic mix.race.Black.Hispanic
## -2.917777e-01 -3.525799e-01
## mix.race.Hispanic.Hispanic
## 9.295829e-01
## Starting unconstrained MCMC...
## Back from unconstrained MCMC.
## Average estimating function values:
## edges degree0
## -182.36719 18.34827
## degree1 mix.identity.msm.msm
## 10.64473 -2777.57688
## mix.identity.msm.tgw mix.identity.tgw.tgw
## -166.77518 -19.94456
## mix.identity.tgw.hrh mix.age_group.18-24.18-24
## -184.53928 79.84271
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 68.83655 77.04911
## mix.age_group.25-34.35-44 mix.age_group.35-44.35-44
## 72.24224 74.58404
## mix.age_group.35-44.45-64 mix.age_group.45-64.45-64
## 78.68190 81.57603
## mix.race.White.White mix.race.White.Black
## 88.06828 101.31663
## mix.race.Black.Black mix.race.White.Hispanic
## 75.61907 88.94817
## mix.race.Black.Hispanic mix.race.Hispanic.Hispanic
## 89.81637 88.51361
## Starting MCMLE Optimization...
## Optimizing with step length 0.806201977527474.
## Using lognormal metric (see control.ergm function).
## Using log-normal approx (no optim)
## The log-likelihood improved by 2.555.
##
## Iteration 4 of at most 500 with parameter:
## edges degree0
## -5.399374e+00 -5.529908e+00
## degree1 mix.identity.msm.msm
## -2.425656e+00 -8.969096e-13
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 8.289876e-12 5.777460e-12
## mix.identity.msm.hrh mix.identity.tgw.hrh
## -Inf -1.528525e-12
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## -Inf -7.380850e+00
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## -7.781753e+00 -5.885590e+00
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## -Inf -7.476746e+00
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## -6.579381e+00 -Inf
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## -Inf -8.132155e+00
## mix.age_group.45-64.45-64 mix.race.White.White
## -7.075352e+00 2.922044e-01
## mix.race.White.Black mix.race.Black.Black
## -7.321493e-02 2.566574e+00
## mix.race.White.Hispanic mix.race.Black.Hispanic
## -3.519415e-01 -3.678120e-01
## mix.race.Hispanic.Hispanic
## 8.785117e-01
## Starting unconstrained MCMC...
## Back from unconstrained MCMC.
## Average estimating function values:
## edges degree0
## -175.078125 9.926398
## degree1 mix.identity.msm.msm
## -7.596476 -2777.576884
## mix.identity.msm.tgw mix.identity.tgw.tgw
## -166.775180 -19.944564
## mix.identity.tgw.hrh mix.age_group.18-24.18-24
## -184.539276 69.997987
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 79.527955 82.323523
## mix.age_group.25-34.35-44 mix.age_group.35-44.35-44
## 71.529352 77.294977
## mix.age_group.35-44.45-64 mix.age_group.45-64.45-64
## 77.877217 81.550644
## mix.race.White.White mix.race.White.Black
## 92.515541 85.259987
## mix.race.Black.Black mix.race.White.Hispanic
## 90.591726 96.351490
## mix.race.Black.Hispanic mix.race.Hispanic.Hispanic
## 91.086875 83.765560
## Starting MCMLE Optimization...
## Optimizing with step length 1.
## Using lognormal metric (see control.ergm function).
## Using log-normal approx (no optim)
## The log-likelihood improved by 0.636.
## Step length converged once. Increasing MCMC sample size.
##
## Iteration 5 of at most 500 with parameter:
## edges degree0
## -5.376750e+00 -5.517903e+00
## degree1 mix.identity.msm.msm
## -2.412497e+00 -8.970059e-13
## mix.identity.msm.tgw mix.identity.tgw.tgw
## 8.289914e-12 5.777451e-12
## mix.identity.msm.hrh mix.identity.tgw.hrh
## -Inf -1.528556e-12
## mix.identity.hrh.hrh mix.age_group.18-24.18-24
## -Inf -7.321663e+00
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## -7.787995e+00 -5.909175e+00
## mix.age_group.18-24.35-44 mix.age_group.25-34.35-44
## -Inf -7.450932e+00
## mix.age_group.35-44.35-44 mix.age_group.18-24.45-64
## -6.569966e+00 -Inf
## mix.age_group.25-34.45-64 mix.age_group.35-44.45-64
## -Inf -8.144472e+00
## mix.age_group.45-64.45-64 mix.race.White.White
## -7.104998e+00 2.746534e-01
## mix.race.White.Black mix.race.Black.Black
## -6.294477e-02 2.583394e+00
## mix.race.White.Hispanic mix.race.Black.Hispanic
## -3.630103e-01 -3.629922e-01
## mix.race.Hispanic.Hispanic
## 8.978471e-01
## Starting unconstrained MCMC...
## Back from unconstrained MCMC.
## Average estimating function values:
## edges degree0
## -164.216797 -4.098993
## degree1 mix.identity.msm.msm
## 1.847860 -2777.576884
## mix.identity.msm.tgw mix.identity.tgw.tgw
## -166.775180 -19.944564
## mix.identity.tgw.hrh mix.age_group.18-24.18-24
## -184.539276 83.866395
## mix.age_group.18-24.25-34 mix.age_group.25-34.25-34
## 79.424195 77.230993
## mix.age_group.25-34.35-44 mix.age_group.35-44.35-44
## 76.732477 81.102106
## mix.age_group.35-44.45-64 mix.age_group.45-64.45-64
## 74.156025 78.450790
## mix.race.White.White mix.race.White.Black
## 90.709388 89.732155
## mix.race.Black.Black mix.race.White.Hispanic
## 94.901541 89.945729
## mix.race.Black.Hispanic mix.race.Hispanic.Hispanic
## 89.303427 95.840267
## Starting MCMLE Optimization...
## Optimizing with step length 1.
## Using lognormal metric (see control.ergm function).
## Using log-normal approx (no optim)
## Starting MCMC s.e. computation.
## Error in solve.default(H, tol = 1e-20) :
## Lapack routine dgesv: system is exactly singular: U[16,16] = 0
## Warning in ergm.MCMCse.lognormal(theta = theta, init = init, statsmatrix =
## statsmatrix0, : Approximate Hessian matrix is singular. Standard errors due
## to MCMC approximation of the likelihood cannot be evaluated. This is likely
## due to insufficient MCMC sample size or highly correlated model terms.
## The log-likelihood improved by 0.3318.
## Step length converged twice. Stopping.
## Finished MCMLE.
## This model was fit using MCMC. To examine model diagnostics and
## check for degeneracy, use the mcmc.diagnostics() function.
theta.casual.form <- casual.fit$coef
theta.casual.form[1] <- theta.casual.form[1] - theta.casual.diss
sim.test <- simulate(casual.net,
formation=formation.casual.net,
dissolution=dissolution,
coef.form=theta.casual.form,
coef.diss=theta.casual.diss,
time.slices=2e4,
#time.slices=1e2,
constraints=constraints,
monitor=~edges+degree(0:5)
)
casual.net.xn <- network.collapse(sim.test, at=20000)
network.size(casual.net.xn)
## [1] 5000
network.edgecount(casual.net.xn)
## [1] 3135
degreedist(casual.net.xn) /network.size(casual.net.xn)
## degree0 degree1 degree2 degree3 degree4
## 869 2139 1849 139 4
## degree0 degree1 degree2 degree3 degree4
## 0.1738 0.4278 0.3698 0.0278 0.0008
mixingmatrix(casual.net.xn, "identity")
## Note: Marginal totals can be misleading
## for undirected mixing matrices.
## HRH MSM TGW
## HRH 0 54 5
## MSM 54 2780 288
## TGW 5 288 8
mixingmatrix(casual.net.xn, "age_group")
## Note: Marginal totals can be misleading
## for undirected mixing matrices.
## 18-24 25-34 35-44 45-64
## 18-24 270 280 0 0
## 25-34 280 687 329 0
## 35-44 0 329 396 315
## 45-64 0 0 315 858
mixingmatrix(casual.net.xn, "race")
## Note: Marginal totals can be misleading
## for undirected mixing matrices.
## Black Hispanic White
## Black 920 184 353
## Hispanic 184 578 487
## White 353 487 613
save.image(file="initial-casual-network.RData")