2

我有这个分位数回归，这些taus:.

taus <-c(.05,.10,.15,.20,.25,.30,.35,.40,.45,.50,.55,.60,.65,.70,.75,.80,.85,.90,.95)
fit1_List<-list()
for(i in 1:length(taus)) {
  fit1_List[[i]]<-rq(foodexp~xx,tau =taus[i],method="br")

}

我想使用stargazer

stargazer(fit1_List, 
          intercept.bottom = FALSE,
          rq.se="nid",ci=TRUE, ci.level=0.90,
          style = "qje",
          title="Quantile Regression Results", align=TRUE,
          dep.var.caption  = "",model.numbers = TRUE
          ,  t.auto = TRUE, p.auto = TRUE,
          font.size="scriptsize", keep.stat="aic",

          dep.var.labels   = "",multicolumn = TRUE

          ,dep.var.labels.include = TRUE)

结果是：

\begin{table}[!htbp] \centering 
  \caption{Quantile Regression Results} 
  \label{} 
\scriptsize 
\begin{tabular}{@{\extracolsep{5pt}}lD{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} D{.}{.}{-3} } 
\\[-1.8ex]\hline 
\hline \\[-1.8ex] 
\\[-1.8ex] & \multicolumn{19}{c}{} \\ 
\\[-1.8ex] & \multicolumn{1}{c}{(1)} & \multicolumn{1}{c}{(2)} & \multicolumn{1}{c}{(3)} & \multicolumn{1}{c}{(4)} & \multicolumn{1}{c}{(5)} & \multicolumn{1}{c}{(6)} & \multicolumn{1}{c}{(7)} & \multicolumn{1}{c}{(8)} & \multicolumn{1}{c}{(9)} & \multicolumn{1}{c}{(10)} & \multicolumn{1}{c}{(11)} & \multicolumn{1}{c}{(12)} & \multicolumn{1}{c}{(13)} & \multicolumn{1}{c}{(14)} & \multicolumn{1}{c}{(15)} & \multicolumn{1}{c}{(16)} & \multicolumn{1}{c}{(17)} & \multicolumn{1}{c}{(18)} & \multicolumn{1}{c}{(19)}\\ 
\hline \\[-1.8ex] 
 Constant & 462.223^{***} & 504.866^{***} & 527.975^{***} & 541.381^{***} & 561.277^{***} & 571.916^{***} & 586.173^{***} & 602.919^{***} & 623.829^{***} & 631.845^{***} & 645.349^{***} & 655.283^{***} & 668.013^{***} & 677.463^{***} & 695.123^{***} & 705.958^{***} & 717.999^{***} & 741.622^{***} & 760.745^{***} \\ 
  & \multicolumn{1}{c}{(431.899$, $492.548)} & \multicolumn{1}{c}{(481.055$, $528.677)} & \multicolumn{1}{c}{(507.211$, $548.740)} & \multicolumn{1}{c}{(523.405$, $559.357)} & \multicolumn{1}{c}{(544.158$, $578.396)} & \multicolumn{1}{c}{(554.395$, $589.437)} & \multicolumn{1}{c}{(567.408$, $604.938)} & \multicolumn{1}{c}{(585.782$, $620.056)} & \multicolumn{1}{c}{(606.506$, $641.152)} & \multicolumn{1}{c}{(614.249$, $649.440)} & \multicolumn{1}{c}{(627.930$, $662.768)} & \multicolumn{1}{c}{(640.077$, $670.489)} & \multicolumn{1}{c}{(651.814$, $684.211)} & \multicolumn{1}{c}{(663.463$, $691.464)} & \multicolumn{1}{c}{(681.002$, $709.245)} & \multicolumn{1}{c}{(690.350$, $721.566)} & \multicolumn{1}{c}{(704.852$, $731.146)} & \multicolumn{1}{c}{(725.196$, $758.047)} & \multicolumn{1}{c}{(749.225$, $772.265)} \\ 
  & & & & & & & & & & & & & & & & & & & \\ 
 xx & 0.343^{***} & 0.402^{***} & 0.424^{***} & 0.447^{***} & 0.474^{***} & 0.481^{***} & 0.489^{***} & 0.510^{***} & 0.552^{***} & 0.560^{***} & 0.566^{***} & 0.586^{***} & 0.605^{***} & 0.609^{***} & 0.644^{***} & 0.660^{***} & 0.678^{***} & 0.686^{***} & 0.709^{***} \\ 
  & \multicolumn{1}{c}{(0.261$, $0.425)} & \multicolumn{1}{c}{(0.336$, $0.468)} & \multicolumn{1}{c}{(0.366$, $0.482)} & \multicolumn{1}{c}{(0.398$, $0.496)} & \multicolumn{1}{c}{(0.426$, $0.522)} & \multicolumn{1}{c}{(0.432$, $0.530)} & \multicolumn{1}{c}{(0.439$, $0.539)} & \multicolumn{1}{c}{(0.462$, $0.558)} & \multicolumn{1}{c}{(0.505$, $0.600)} & \multicolumn{1}{c}{(0.514$, $0.607)} & \multicolumn{1}{c}{(0.519$, $0.612)} & \multicolumn{1}{c}{(0.545$, $0.627)} & \multicolumn{1}{c}{(0.565$, $0.644)} & \multicolumn{1}{c}{(0.573$, $0.645)} & \multicolumn{1}{c}{(0.606$, $0.682)} & \multicolumn{1}{c}{(0.616$, $0.703)} & \multicolumn{1}{c}{(0.640$, $0.715)} & \multicolumn{1}{c}{(0.639$, $0.733)} & \multicolumn{1}{c}{(0.680$, $0.739)} \\ 
  & & & & & & & & & & & & & & & & & & & \\ 
\hline 
\hline \\[-1.8ex] 
\textit{Notes:} & \multicolumn{19}{r}{$^{***}$Significant at the 1 percent level.} \\ 
 & \multicolumn{19}{r}{$^{**}$Significant at the 5 percent level.} \\ 
 & \multicolumn{19}{r}{$^{*}$Significant at the 10 percent level.} \\ 
\end{tabular} 
\end{table}

如您所见，格式是水平的。如何垂直放置？也就是说，最后我将只有两列。

最重要的是，保持星号的重要性。

4

1 回答 1

0

我认为没有一种简单的方法可以做到这一点stargazer。作为stargazer用户，我创建了几个用户定义的函数来概括这个包的使用，在额外的编码之后，将产生接近你想要的结果的东西。

例如，

taus <-c(.05,.10,.15,.20,.25,.30,.35,.40,.45,.50,.55,.60,.65,.70,.75,.80,.85,.90,.95)
fit1_List<-list()

#First I save all the coef and test statistics from the specifications.  Notice I use the iris data set since you did not provide data.

MYcoef <-  matrix(NA, nrow = length(taus), ncol = 2)
MYteststat <- matrix(NA, nrow = length(taus), ncol = 2)
for(i in 1:length(taus)) {
  res <- rq(iris$Sepal.Length ~ iris$Sepal.Width, tau = taus[i], method = "br")
  MYcoef[i, ] <- res$coef
  MYteststat[i, ] <- summary(res, se = "nid")$coef[, 3]  
}

#Then utilize my user defined function, myStars, to star all the coefficients.  This little function is really helpful and I use it all the time when reporting results.  

myStars <- function(testStat, cval, sig_dig, coef = FALSE, coefValue = NULL){
  testStat<- round(testStat, sig_dig)

  if(!is.matrix(testStat))
    test <- matrix(testStat, length(cval), 1)

  if(!is.matrix(cval))
    cval <- matrix(cval, 1, length(cval))

  test <- matrix(rep(testStat, ncol(cval)),nrow(cval), ncol(cval))

  starCount <- rowSums(abs(test) > abs(cval))

  stars <- sapply(starCount, function(x) paste(rep("*", x), collapse = ""))

  if(coef)
    paste0(round(coefValue, sig_dig), stars)
  else
    paste0(testStat, stars)
}



#I don't know much about QR, however, I suspect the critical values are incorrect below. 

MYcoefStars <-  matrix(NA, nrow = length(taus), ncol = 2)
for(i in 1:nrow(MYcoef)){
  for(j in 1:ncol(MYcoef)){
    MYcoefStars[i,j] <- myStars(MYteststat[i, j], c(1.64, 1.96, 2.33), 2, coef = TRUE, coefValue = MYcoef[i, j])
  }
}

#Finally, I've starred all the coefficients, however, stargazer is going to interpret the asterisk as a undesirable latex character.  The below user defined function fixStar, as its names suggests, fixes this.  It is essentially a wrapper for stargazer, so you can pass it any stargazer argument you like.

fixStar <- function(x,...){
  tbl <- capture.output(stargazer(x, ...))
  undo <- gsub("\\\\textasteriskcentered", "*", tbl)
  restar <- gsub("* * *", "${}^{***}$", undo, fixed = TRUE)
  restar <- gsub("* *", "${}^{**}$", restar, fixed = TRUE)
  restar <- gsub("* ", "${}^{*}$", restar, fixed = TRUE)
  cat(restar, sep="\n")
}

fixStar(MYcoefStars)

这应该能让你大部分时间到达那里。表格的格式需要工作，并且可能需要在乳胶中进行一些编码。话虽如此，我认为上述两个用户定义的功能可能对其他stargazer用户有用。

于 2016-07-08T21:39:54.433 回答