0

我对 R 非常缺乏经验，但据我所知，该Synth包应该创建一个合成投影“控制组”。我正在尝试创建一个综合预测，如果该州没有废除死刑，该州的凶杀率将会是多少 - 对于上下文，我正在查看 1996-2014 年的纽约，死刑在2007 年在那里，所以我正在寻找 2007-2014 年的预测。Synth似乎讨厌我坚持那个状态的数据，它一直在要求控制数据......我认为重点是创建一个综合控制。

下面是一些尝试。任何有关如何设置此 data.prep 以便为我提供 2007-2014 年凶杀率的综合预测的见解将不胜感激。不确定这是否会脱离上下文有任何意义，但这里是：

NewYork$STATE2 <- as.character(NewYork$STATE)
dataprep.out <-
dataprep(foo = NewYork, predictors = c("YEAR", "DEATH.PENALTY"), 
         predictors.op = "mean", dependent = "HOMICIDE", 
         unit.variable = "DEATH.PENALTY", time.variable = "YEAR",
         treatment.identifier = 0, controls.identifier = 1,
         time.predictors.prior = c(1996:2006), time.optimize.ssr = c(1996:2007),
          unit.names.variable = "STATE2", time.plot = 1996:2014)

数据准备错误（foo = NewYork, predictors = c("YEAR", "DEATH.PENALTY"), :
请至少指定两个控制单元

dataprep.out <- dataprep(foo = NewYork, predictors = c("YEAR", "DEATH.PENALTY"),
                         predictors.op = "mean", dependent = "HOMICIDE",
                         unit.variable = "DEATH.PENALTY", time.variable = "YEAR",
                        treatment.identifier = 0, controls.identifier = c(1,1),
                        time.predictors.prior = c(1996:2006),
                        time.optimize.ssr = c(1996:2007),
                        unit.names.variable = "STATE2",
                        time.plot = 1996:2014)

dataprep(foo = NewYork, predictors = c("YEAR", "DEATH.PENALTY"), 中的错误：controls.identifier 中的控制单元重复

NewYork$DEATH.PENALTY2 <- ifelse(NewYork$YEAR<2000,2,NewYork$DEATH.PENALTY)
dataprep.out<-
    dataprep(foo = NewYork, predictors = c("YEAR", "DEATH.PENALTY2"),
    predictors.op = "mean", dependent = "HOMICIDE", unit.variable = "DEATH.PENALTY2",
    time.variable = "YEAR", treatment.identifier = 0, controls.identifier = c(1,2),
    time.predictors.prior = c(1996:2006), time.optimize.ssr = c(1996:2007),
    unit.names.variable = "STATE2", time.plot = 1996:2014)

数据准备中的错误（foo = NewYork，predictors = c("YEAR", DEATH.PENALTY2")，：控制中的处理单位**

4

0 回答 0