使用ggplot2绘制缺失值

Question

我正在尝试使用下面的函数绘制缺失值。我收到此错误消息：

1101b

101b

有什么想法吗？

Answer 1

ggplot中的x和y未在您的函数中指定。我将其更改为以下内容：

ggplot_missing <- function(data){
  df2 <- data %>% is.na %>% melt 

  ggplot(df2, aes(Var2, Var1, fill=value)) + 
    geom_raster() + 
    scale_fill_grey(name="", labels=c("Present", "Missing")) +
    theme_minimal() + 
    theme(axis.text.x  = element_text(angle=45, vjust=0.5)) + 
    labs(x = "Variables in Dataset",
         y = "Rows / observations")
}

测试数据：

df <- iris
set.seed(4)
df[sample(nrow(df), 20), 2] <- NA
df[sample(nrow(df), 30), 3] <- NA
df[sample(nrow(df), 15), 4] <- NA

ggplot_missing(df)

Answer 2

OP的问题略有不同。如果要在另一个（因子）变量的不同级别可视化每个变量的缺失数据模式......

ggplot_missing2 <- function(data, xvar, yvars) {
  # xvar should be a factor variable for this to work
  require(ggplot2)
  require(reshape2)
  newvar = "variable"
  newval = "value"
  dl <- melt(data, id.vars = xvar, measure.vars=yvars, variable.name=newvar, value.name = newval)
  dl <- dcast(dl, formula = as.formula(paste0(newvar,"~",xvar)),
              fun.aggregate = function(x) sum(is.na(x)))
  dl <- melt(dl, id.vars=newvar, variable.name=xvar, value.name=newval)
  ggplot(dl, aes_string(x=xvar, y=newvar)) + 
    geom_tile(aes_string(fill=newval), color="white") +
    geom_text(aes_string(label=newval)) + 
    scale_fill_continuous("Missing (N)", low="gray", high="cornflowerblue") +
    labs(title="Missing Data Pattern")
}

测试数据：

df <- iris
set.seed(4)
df[sample(nrow(df), 20), 2] <- NA
df[sample(nrow(df), 30), 3] <- NA
df[sample(nrow(df), 15), 4] <- NA

ggplot_missing2(df)