Question

我在ggplot2库中遇到ggsave（）问题。我编写了一个函数，我将参数传递给，并且应该生成然后使用ggsave（）保存结果。

以下是重现错误的一些示例数据和代码：

example.df.1 <- data.frame(matrix(1:100, nrow = 20, ncol = 5))
colnames(example.df.1) <- c("var1", "var2", "var3", "var4", "var5")
rownames(example.df.1) <- c("A", "B", "C", "D", "E", "F", "G", "H", 
                            "I", "J", "K", "L", "M", "N", "O", "P",
                            "Q", "R", "S", "T")

example.df.2 <- data.frame(matrix(ncol = 2, nrow = 24))
example.df.2[,1] <- c("A", "B", "C", "D", "E", "F", "G", "H", 
                            "I", "J", "K", "L", "M", "N", "O", "P",
                            "Q", "R", "S", "T", "U", "V", "W", "X")

example.df.2[,2] <- rnorm(24, 10, 2)                            

problematic_func <- function(data1, col, title, var, data2)  {
    # only include rows without missing values
    loc1 <- subset(data1, rowSums(is.na(data1)) == 0)

    loc1 <- cbind(loc1, rank(-as.data.frame(loc1[,1]), ties.method = "first"))

    # reduce data2 to only those rows that correspond to rows in data1
    loc2 <- data2[data2[,1] %in% rownames(loc1),]

    # order loc2
    loc2.ordered <- loc2[order(loc2[,1]),]

    # correlation between loc1 and loc2.ordered
    corr <- cor(loc1[,1], loc2.ordered[,2])

    # creating the plot
    i <- ggplot(loc1, aes_q(x = loc1[,1], y = loc2.ordered))
    i <- i + geom_point(colour = col, size = 4)
    i <- i + ggtitle(title)
    i <- i + xlab(var)
    i <- i + ylab("y-axis")
    i <- i + coord_cartesian(xlim = c(0, max(loc1[,1])),
        ylim = c(0, max(loc2.ordered[,2])*1.2))
    i <- i + annotate("text", x = max(loc1[,1])*.5, y = 1,
        label = paste("Correlation coef: ", as.character(corr)), size = 3)

    # saving the plot - this is where the error occurs according
    # to the debugger   
    ggsave(filename = paste("my_example_plot_", var, ".png", sep = ""),
        plot = i, device = png, width = 625, height = 625, limitsize = FALSE)
}

for (i in 1:ncol(example.df.1)) {
    sv <- as.data.frame(example.df.1[,i])
    rownames(sv) <- rownames(example.df.1)
    problematic_func(sv, "orange", colnames(example.df.1[i]),
        colnames(example.df.1[i]), data2 = example.df.2)
}

编辑：抱歉，我忘了添加错误：

FUN中的错误（X [[2L]]，...）：未知输入：data.frame

Answer 1

我自己发现了这个错误。错误不在ggsave（）函数中，而是在我首先创建绘图的ggplot（）函数中。正确的代码必须是：

i <- ggplot(loc1, aes_q(x = loc1[,1], y = loc2.ordered[,2]))

不同之处在于，y审美需要传递一个loc2.ordered列，而不是整个数据帧。

ggave（）错误：输入R未知

1 个答案: