R 在HydeNet中指定决策节点的自定义模型

我正在使用

HydeNet

软件包构建贝叶斯决策网络：

市场节点和调查节点分别具有自定义概率和cpt性质。其思想是将发现节点定义为具有以下逻辑的决策节点：代理根据调查节点信息计算其效用，然后做出是否在市场上发现公司的决策。在这里，我分别计算了ExpUtility值，然后将它们粘贴到determ节点：

# Found node (how to put down an optimization task on it?)
founding <- data.frame(Found = as.factor(c(rep(0,3), rep(1,3))),
                       Survey = as.factor(rep(c(1,2,3),2)),
                       EV = as.numeric(c(NA,NA,NA,NA,NA,NA)))

expvalue = function(founding_option, survey_option){
  df = outcomes %>%  
    left_join(market_status, by = "Market") %>% 
    inner_join(survey_cond, by = "Market") %>% 
    filter(Found == founding_option) %>% 
    filter(Survey == survey_option) %>% 
    mutate(expexted = as.numeric(as.character(Utility))*prob_real*prob_surv)
  ev = sum(df$expexted)
  return(ev)
}

for (i in 1:nrow(founding)) {
  founding$EV[i] = expvalue(founding$Found[i], founding$Survey[i])
}

net <- setNode(net, Found, nodeType = "determ", define=fromFormula(),
              nodeFormula = Found~ifelse(Survey == "1", 0, 1))
net <- setDecisionNodes(net, Found)

#找到节点（如何在其上放置优化任务？）
创始%
左加入（市场状态，由=“市场”）%>%
内部加入（调查条件，按=“市场”）%>%
过滤器（已找到==U选项）%>%
过滤器（调查==调查选项）%>%
mutate（expexted=as.numeric（as.character（Utility））*prob_real*prob_surv）
ev=总额（df$expexted）
返回（ev）
}
适用于（i/1:nrow（创始））{
创始$EV[i]=expvalue（创始$Found[i]，创始$Survey[i]）
}
net我不知道HydeNet，但使用Pyagram，您可以很容易地做到这一点：

infdiag=gum.fastd（“市场{poor | medium | perspective}->Survey{poor | medium | perspective}->*发现{nOK | OK}->$Utility[3]我真的不明白为什么P（Survey | Market）中的值是错误的：这个参数向我们展示了某些调查值的概率（s1、s2、s3）在每个市场状态下。每个市场状态的总和是1，所以它是正确的，不是吗？此外，感谢您的解决方案，但由于大量的软件包和分析，我更喜欢使用R。当然没问题，我只是想展示一个看起来更简单的解决方案（对我来说）.关于你的参数，你是对的..我的错：我直接把这个列表0.6,0.3,0.1,0.3,0.4,0.4,0.1,0.1,0.3,0.5作为CPT…但是对于Pyagram，它应该被转置：0.6,0.3,0.1,0.3,0.4,0.3,0.1,0.5（这样更容易检查总和；-）。
library(dplyr)
library(HydeNet)

# 1 - poor, 2 - medium, 3 - perspective
market_status = data.frame(Market = as.factor(c(1,2,3)),
                           prob_real = as.numeric(c(0.5, 0.3, 0.2)))

# the probability of defining the market status
survey_cond = data.frame(Market = as.factor(rep(c(1,2,3),3)),
                         Survey = as.factor(c(rep(1,3), rep(2,3), rep(3,3))),
                         prob_surv = as.numeric(c(0.6, 0.3, 0.1, 0.3, 0.4, 0.4, 0.1, 0.3, 0.5)))


# final utilities
outcomes = data.frame(Found = as.factor(c(rep(0,3), rep(1,3))),
                      Market = as.factor(rep(c(1,2,3),2)),
                      Utility = as.factor(c(0,0,0,-7,5,20)))


net <- HydeNetwork(~ Utility | Found*Market
                   + Found | Survey
                   + Survey | Market)

plot(net)

# Market node 
net <- setNode(net, Market, nodeType="dcat",  pi=vectorProbs(p=market_status$prob_real, Market), 
               factorLevels = c("1","2","3"))

# Survey node 
survey.cpt <- cpt(Survey ~ Market,
                  data = survey_cond,
                  wt = survey_cond$prob_surv)
net <- setNodeModels(net, survey.cpt)

# Found node (how to put down an optimization task on it?)
founding <- data.frame(Found = as.factor(c(rep(0,3), rep(1,3))),
                       Survey = as.factor(rep(c(1,2,3),2)),
                       EV = as.numeric(c(NA,NA,NA,NA,NA,NA)))

expvalue = function(founding_option, survey_option){
  df = outcomes %>%  
    left_join(market_status, by = "Market") %>% 
    inner_join(survey_cond, by = "Market") %>% 
    filter(Found == founding_option) %>% 
    filter(Survey == survey_option) %>% 
    mutate(expexted = as.numeric(as.character(Utility))*prob_real*prob_surv)
  ev = sum(df$expexted)
  return(ev)
}

for (i in 1:nrow(founding)) {
  founding$EV[i] = expvalue(founding$Found[i], founding$Survey[i])
}

net <- setNode(net, Found, nodeType = "determ", define=fromFormula(),
              nodeFormula = Found~ifelse(Survey == "1", 0, 1))
net <- setDecisionNodes(net, Found)

# Utility node
net <- setNode(net, Utility, "determ", define=fromFormula(),
               nodeFormula = Utility ~ ifelse(Found == 0, 0,
                                              ifelse(Market == 1, -7, 
                                                     ifelse(Market == 2, 5, 20))))
net <- setUtilityNodes(net, Utility)

plot(net)

trackedVars <- c("Market","Survey","Found","Utility")
evidence <- NULL
compiledNet <- compileJagsModel(net, data = evidence,
                                n.chains = 3,
                                n.adapt = 5000)
post <- HydeSim(compiledNet,
                variable.names = trackedVars,
                n.iter=10000)

dplyr::sample_n(post, 20)
table(post$Utility)
mean(post$Utility)


      ||  Found            |
Survey||nOK      |OK       |
------||---------|---------|
poor  || 1.0000  | 0.0000  |
medium|| 0.0000  | 1.0000  |
perspe|| 0.0000  | 1.0000  |