R ggplot2尝试制作重复叠加面积图；获取不一致的错误

作为一个更大脚本的一部分，我必须制作70个堆叠面积图，每年1个，然后将它们写入pdf。我只是偶尔有问题。我把它设置成一个循环，有时在前四年中断，有时完成一半，有时几乎结束

当我跨过循环，不管它在哪一年中断（通过取消注释ThisYear和window行），我都可以让图形在窗口中显示得很好。有时候。有时我会两次绘制同一年的图表，第一次有效，第二次无效。有时反之亦然

发生错误时，可能是“错误：结果必须全部是原子的，或所有数据帧”，或“错误：参数意味着不同的行数：14641458”，或“错误：不能在CHARSXP上有属性”，但我无法可靠地使相同的错误再次发生。它们发生在打印（p）时，但我不知道“p”是什么让它崩溃的

我使用的数据（下面冒充为“BlueData”、“GreenData”、“YellowData”和“SalmonData”）都是双倍的，我已经检查并重新检查过了。没有NAs。事实上，当我伪造数据在这里发布这篇文章时，我以为我有什么东西，因为它没有崩溃。然后我运行了几次，它也开始以同样不一致的方式抛出同样的错误

编辑：我可能缺少一些“require”语句。。。我很快就会修好的

require(lubridate)

TestYear = 1929

BlueLimit = 110

### creating fake data to mimic what my data is like

BlueData <- rnorm(70*365+14, mean = 100, sd = 25)
BlueData <- ifelse(BlueData > BlueLimit, BlueLimit, BlueData)
YellowData <- rnorm(70*365+14, mean = 15, sd = 2)
SalmonData <- rnorm(70*365+14, mean = 20, sd = 3)
GreenData <- rnorm(70*365+14, mean = 5, sd = 1)
BigData.df <- as.data.frame(cbind(BlueData,YellowData,SalmonData,GreenData))
BigData.df$Date <- seq(from = ISOdate(TestYear - 1, 10, 1), length.out = nrow(BigData.df), by = "day")
BigData.df$MainData <- BigData.df$BlueData + BigData.df$GreenData + BigData.df$YellowData + BigData.df$SalmonData
BigData.df$LimitData <- BlueLimit

### long complicated reasons, but no February 29ths allowed in the dataset...

BigData.df <- subset(BigData.df, (month(BigData.df$Date) == 2)*(day(BigData.df$Date) == 29) == 0)

### okay, that should do it for mocking up data...

pdfFilename <- "C:\\TestFolder\\TestPDF.pdf"


plot_colors_aggregate = c("gold","forestgreen","lightsalmon","dodgerblue")


    ### establish appropriate scale for top of graph and gridlines

max_y <- max(BigData.df$MainData)
max_y <- ceiling(max_y/50)*50
if (max_y <= 200) {
    gridline_multiplier = 20 
    } else {
    gridline_multiplier = 50
}




pdf(file = pdfFilename, height=8, width = 11, onefile = TRUE)


for (ThisYear in 1929:1998){

    #GraphYear = TestYear
    #windows(height = 8, width = 12)

    cat("\t\tNow working on",ThisYear)

    page_title = paste("This Graph is for",ThisYear)

    ##### CREATE SUBSET OF BIGDATA.DF WITH JUST THIS GRAPH YEAR

    GraphData.df <- subset(BigData.df, year(BigData.df$Date) == ThisYear)


    ##### CREATE SMALLER DF WITH JUST DATA TO BE GRAPHED

    df4plot <- data.frame(
                    AllColors = c(GraphData.df$BlueData, 
                            GraphData.df$SalmonData, 
                            GraphData.df$GreenData, 
                            GraphData.df$YellowData),
                    date = rep(GraphData.df$Date, 4),
                    DataType = rep(c("Blue Stuff", "Salmon Stuff", "Green Stuff", "Yellow Stuff"), each = 365)
                    )


    df4LimitLine <- data.frame(
                    BlueLimit = GraphData.df$LimitData,
                    date = GraphData.df$Date
                    )



    df4plot$DataType <- factor(df4plot$DataType, 
                levels = c("Yellow Stuff", "Green Stuff","Salmon Stuff","Blue Stuff"))


    #PeriodBreaks <- GraphData.df$Date[FirstOfTheMonth.v]   



    scale_color_manual(values=plot_colors_aggregate)


    p <- ggplot(data = df4plot, aes(date, AllColors)) + 
        scale_fill_manual(values = plot_colors_aggregate) + 
        scale_color_manual(values = plot_colors_aggregate)+ 
        scale_y_continuous(limits=c(0,max_y), breaks=seq(0,max_y,gridline_multiplier*2))


    p <- p + 
        geom_area(aes(colour = DataType, fill = DataType), position="stack")  +
        theme(axis.text.x = element_text(angle = -45, hjust = 0, vjust=1)) + 
        theme(legend.position=c(.7, .7), legend.title=element_blank(), legend.justification = "left", legend.background = element_rect(fill="white")) +
        theme(panel.background = element_rect(), panel.grid.major = element_line(colour = "darkgrey", linetype = "dotted"), panel.grid.minor = element_line(colour="lightgrey", linetype = "dotted")) +
        theme(plot.margin = unit(c(1,1,5,1),"line"))

    p <- p + geom_line(data=df4LimitLine, aes(x=date,y=BlueLimit, linetype = "Limit of Blue Data") ,colour="darkblue") +
        theme(legend.key = element_rect(fill="white", linetype = 0)) +
        scale_linetype_manual(name = "Linetype", values = "dashed") +
        annotate("text", x=df4LimitLine$date[floor(nrow(df4LimitLine)/2)], y=max_y, label=page_title) 



    print(p)

    grid.text("Blah blah blah", x=unit(.95,"npc"), y=unit(.1,"npc"), hjust=1, gp=gpar(cex = .8))
    grid.text("Yadda yadda yadda", x=unit(.95,"npc"),y=unit(.08,"npc"), hjust = 1, gp=gpar(cex = .8))

    cat(ThisYear,"now finished.\n")
}


dev.off()

require（润滑油）
测试年=1929年
蓝极限=110
###创建虚假数据以模拟我的数据
BlueDataI解决了不一致的错误和错误：结果必须全部是原子的，或者通过恢复到r3.1.1，所有的数据帧。3.2.0似乎不稳定。
关于不同行数的消息可能是由您创建
df4plot
的方式造成的-您可能会在一年中没有365天数据的任何时候收到此错误。考虑使用代码从<代码> RESHAPE2或<代码> TyDyr < /代码>对数据集进行整形，以避免此问题。我无法复制任何其他错误消息。