在对数据进行一些更新后，图例在facet_grid图中发生了变化

Question

我有一个ggplot barplot与facet_grid一直工作到今天。我不知道为什么传说与以前不一样。现在没有4状态，我有很多，如图所示。

用于创建数据帧的代码：

df <- structure(list(Business = c("B", "B", "B", "K", "K", "K", "K", 
"N", "N", "N", "N", "N", "O", "O", "P", "P", "P", "P", "P", "P", 
"W", "W", "W"), Status = structure(c(1L, 2L, 3L, 1L, 2L, 3L, 
4L, 3L, 1L, 2L, 3L, 4L, 2L, 3L, 1L, 2L, 3L, 4L, 2L, 3L, 1L, 2L, 
3L), .Label = c("Cancelled", "Closed", "In SAP", "Open"), class = "factor"), 
    `Status Count` = c(5L, 22L, 174L, 40L, 65L, 551L, 4L, 36L, 
    2L, 4L, 43L, 5L, 10L, 4L, 22L, 132L, 394L, 6L, 2L, 5L, 6L, 
    66L, 106L), Percentage = c(2.49, 11, 87, 6.06, 10, 83, 1, 
    100, 3.7, 7, 80, 9, 71, 29, 3.97, 24, 71, 1, 29, 71, 3.37, 
    37, 60), tot = c(75L, 301L, 1313L, 75L, 301L, 1313L, 15L, 
    1313L, 75L, 301L, 1313L, 15L, 301L, 1313L, 75L, 301L, 1313L, 
    15L, 301L, 1313L, 75L, 301L, 1313L), business_per = c(4, 
    18, 77, 4, 18, 77, 0.88, 1, 77, 4, 18, 0.88, 77, 1, 18, 77, 
    4, 0.88, 18, 77, 1, 18, 77)), .Names = c("Business", "Status", 
"Status Count", "Percentage", "tot", "business_per"), row.names = c(NA, 
-23L), class = c("tbl_df", "tbl", "data.frame"))


#create the legend
MStatus=paste(df$Status,"(",df$tot,")",df$business_per,"%")

情节：

ggplot(df, aes(fill=`MStatus`, y=`Status Count`, x=Status, color=factor(Status))) + 
  geom_bar(position=position_dodge(width = 2), stat="identity") +
  facet_grid(~Business, labeller = labeller(Business=label_wrap_gen(12))) +
  geom_text(size=4, aes(label=paste(Percentage,"%"), vjust = -0.5)) +
  scale_color_discrete(guide=FALSE) + # removes the legend for factor(Status)
  scale_x_discrete(breaks = NULL) + #turn off vertical grid lines and removed the axis x label
  theme(axis.text.x = element_text(angle=90, hjust=1))

这是我得到的情节和传奇：

这是我期待的传奇（数字不正确，风格很重要）

编辑1：

我按照相同的步骤创建以下数据框，如果正常则创建图例。我在这里看到的唯一区别只是记录的数量，这不应该导致问题。这是函数中的错误吗？

df <- structure(list(Business = c("B", "B", "B", "K", "K", "K", "K", 
"N", "N", "N", "N", "N", "O", "O", "P", "P", "P", "P", "P", "P", 
"W", "W", "W"), Status = structure(c(1L, 2L, 3L, 1L, 2L, 3L, 
4L, 3L, 1L, 2L, 3L, 4L, 2L, 3L, 1L, 2L, 3L, 4L, 2L, 3L, 1L, 2L, 
3L), .Label = c("Cancelled", "Closed", "In SAP", "Open"), class = "factor"), 
    `Status Count` = c(5L, 22L, 174L, 40L, 65L, 551L, 4L, 36L, 
    2L, 4L, 43L, 5L, 10L, 4L, 22L, 132L, 394L, 6L, 2L, 5L, 6L, 
    66L, 106L), Percentage = c(2.49, 11, 87, 6.06, 10, 83, 1, 
    100, 3.7, 7, 80, 9, 71, 29, 3.97, 24, 71, 1, 29, 71, 3.37, 
    37, 60), tot = c(75L, 301L, 1313L, 75L, 301L, 1313L, 15L, 
    1313L, 75L, 301L, 1313L, 15L, 301L, 1313L, 75L, 301L, 1313L, 
    15L, 301L, 1313L, 75L, 301L, 1313L), business_per = c(4, 
    18, 77, 4, 18, 77, 0.88, 1, 77, 4, 18, 0.88, 77, 1, 18, 77, 
    4, 0.88, 18, 77, 1, 18, 77)), .Names = c("Business", "Status", 
"Status Count", "Percentage", "tot", "business_per"), row.names = c(NA, 
-23L), class = c("tbl_df", "tbl", "data.frame"))



#create the legend
MStatus=paste(df$Status,"(",df$tot,")",df$business_per,"%")

这是使用新数据框创建的图表：

Answer 1

最后我弄清楚问题是什么：

在df，business_per用于构建传说，每个状态需要相同的数字。罪魁祸首是以下线：

df$business_per[df$business_per>=1] <- round(df$business_per,0)

应该在哪里：

df$business_per[df$business_per>=1] <- round(df$business_per[df$business_per>=1],0)