• • bounty将在5小时后过期**。回答此问题可获得+50声望奖励。rudgus51998希望引起更多人关注此问题。

我将glmmTMB和pscl中的zeroinfl应用于同一个数据集，得到了条件部分相同的系数，但二进制部分的系数有些不同，您知道造成差异的潜在因素吗？
谢谢!
下面是zeroinfl的结果：

# > summary(pscl.res)
# 
# Call:
#   zeroinfl(formula = outside_treatment ~ group + baseline.risk + Age.group + 
#              offset(log(Follow.up)) | group, data = final, dist = "negbin", 
#            link = "logit", trace = TRUE)
# 
# Pearson residuals:
#     Min      1Q  Median      3Q     Max 
# -1.0332 -0.7003 -0.4041  0.1675 12.5728 
# Count model coefficients (negbin with log link):
#                Estimate Std. Error z value Pr(>|z|)    
# (Intercept)    -2.39168    0.13282 -18.007  < 2e-16 ***
# group1         -0.22475    0.09833  -2.286   0.0223 *  
# baseline.risk2 -0.20385    0.12947  -1.575   0.1154    
# baseline.risk3 -0.67367    0.12422  -5.423 5.85e-08 ***
# Age.group1     -0.65774    0.11554  -5.693 1.25e-08 ***
# Log(theta)      0.12366    0.07302   1.693   0.0904 .
# Zero-inflation model coefficients (binomial with logit link):
#             Estimate Std. Error z value Pr(>|z|)
# (Intercept)  -16.564    389.991  -0.042    0.966
# group1        -2.743   1909.909  -0.001    0.999
# ---
# Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 
# 
# Theta = 1.1316 
# Number of iterations in BFGS optimization: 35 
# Log-likelihood: -1531 on 8 Df

这是glmmTBM的结果：

# > summary(true.zinb)
# Family: nbinom2  ( log )
# Formula:          
# outside_treatment ~ group + baseline.risk + Age.group + offset(log(Follow.up))
# Zero inflation:                     ~1 + group
# Data: final
# 
# AIC      BIC   logLik deviance df.resid 
# 3078.1   3110.5  -1531.1   3062.1      414 
# 
# 
# Dispersion parameter for nbinom2 family (): 1.13 
# 
# Conditional model:
#                Estimate Std. Error z value Pr(>|z|)    
# (Intercept)    -2.39168    0.13282 -18.007  < 2e-16 ***
# group1         -0.22475    0.09833  -2.286   0.0223 *  
# baseline.risk2 -0.20385    0.12947  -1.575   0.1154    
# baseline.risk3 -0.67367    0.12422  -5.423 5.85e-08 ***
# Age.group1     -0.65774    0.11554  -5.693 1.25e-08 ***
#   ---
# Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
# 
# Zero-inflation model:
#              Estimate Std. Error z value Pr(>|z|)
# (Intercept)  -20.8282  3289.0961  -0.006    0.995
# group1         0.4139  4626.1083   0.000    1.000