2

我有一个矩阵，它显示了我用于某些线性组合的变量的分类值和相应的 r.squared 值。由于用于线性组合的变量数量在 1 到 4 之间变化，因此存在多个 NA。

我使用 ggplot2 来获得这个输出，这非常接近我想要的： Rplot1.png

但是，我不希望显示 NA，并且我无法弄清楚如何geom_text()显示所有内容，或者如果我尝试类似的操作给我一个错误，geom_text(aes(label = sprintf("%1.0f",value[!is.na(value])))因为在这种情况下长度是错误的。

此外，我希望以不同的精度（2 个小数位）和不同的颜色代码格式化最后一列。

我的数据：

col1 <- c(3, 12, 12,  4,  2,  4,  6,  4, 15, 13,  1,  4, 13, 13,  4,  4,  6, 13,  1,  4)
col2 <- c(14, 29, 33, NA,  3, 14, 14, NA, 18, 18,  3, NA, 18, 18,  5, NA, 14, 18,  3, NA)
col3 <- c(22, 33, NA, NA,  6, 23, NA, NA, 25, 25, NA, NA, 24, 25, NA, NA, 21, 25, NA, NA)
col4 <- c(25, NA, NA, NA,  9, NA, NA, NA, 31, NA, NA, NA, 25, NA, NA, NA, 27, NA, NA, NA)
col5 <- c(0.99879174731700204326, 0.99026853951234095064, 0.97577109510739901133, 0.94129388844778405510, 0.99280352772128999117, 0.96836063265002103595, 0.93791976583184200056,
      0.84138009684215697703, 0.98856063354420098399, 0.97136223282556599656, 0.79804189507803902082, 0.64147943288862196987, 0.99825754796980703532, 0.96893762747413503256,
      0.87923096638698294836, 0.78073080902317304464, 0.99370888114350197640, 0.95233511051330899111, 0.80470423703937299376, 0.62955867989309499677)

plotMat2 <- data.frame(cbind(col1,col2,col3,col4,col5),
                            row.names=c("data.max4","data.max3","data.max2","data.max1","data.3QT4","data.3QT3","data.3QT2","data.3QT1","data.median4","data.median3","data.median2",
                              "data.median1","data.mean4","data.mean3","data.mean2","data.mean1","data.1QT4","data.1QT3","data.1QT2","data.1QT1"))
names(plotMat2) <- c("SU" ,  "SU.1", "SU.2", "SU.3", "rsq" )

我的代码：

library(reshape)
cor_dat2 <- melt(as.matrix(plotMat2))
names(cor_dat2) <- c('Var1','Var2','value')
cor_dat2$Var1 <- factor(row.names(plotMat2), levels=rev(row.names(plotMat2)))
cor_dat2$Var2 <- factor(cor_dat2$Var2, levels=colnames(plotMat2))


library(ggplot2)
p <- ggplot(data =  cor_dat2, aes(x = Var2, y = Var1)) +
  geom_tile(aes(fill = value), colour = "black") +
  geom_text(aes(label = sprintf("%1.0f",value))) +
  scale_fill_gradient(low = "white", high = "red")
print(p)

任何帮助表示赞赏，在此先感谢，埃里克

4

1 回答 1

8

只需将其替换为geom_text：

geom_text(aes(label = ifelse(is.na(value), "", sprintf("%1.0f",value)))) +

在此处输入图像描述

于 2013-04-06T10:17:44.440 回答