Model: fit_precheck_edges_persist
Speed mode: strict_filtered_core_mechanism_transition_reduced_to_chinese_paper_variables
Mode note: strict data filter + patent-size cap(<=5) + top-node cap + matrix-based NA-dyad networks + NetSeries(NA.impute='0') + symmetrical formation/persistence specification + main structural path(edges, gwesp) + alternative structural path(edges, gwdsp) + lagged node capability(evcent, tech_size) + static dyadic covariates(proximity, history) + period descriptive statistics + transition-specific full-model coefficients + true single-core execution (no cluster) + GOF only for the mechanisms actually retained in the final model
Formula used: series ~ Form(~ edges) + Persist(~ edges)
Fixed roster: YES
Structural zero implementation: NA dyads + NetSeries(NA.impute='0')

Call:
ergm(formula = series ~ Form(~edges) + Persist(~edges), control = make_control(burnin = 25000, 
    interval = 2000, samplesize = 256, maxit = 80), verbose = TRUE)

Maximum Likelihood Results:

                 Estimate Std. Error MCMC % z value Pr(>|z|)    
Form(1)~edges    -2.82838    0.03471      0  -81.48   <1e-04 ***
Persist(1)~edges  1.45276    0.07647      0   19.00   <1e-04 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

     Null Deviance: 23377  on 16863  degrees of freedom
 Residual Deviance:  7862  on 16861  degrees of freedom
 
AIC: 7866  BIC: 7882  (Smaller is better. MC Std. Err. = 0)
