R时间序列间隙填充用于绘制类型='b'

Question

我有一个 .csv 文件，其中包含四列（NAME、ID、YEAR、VALUE，请参见下面的示例），并希望使用 plot('YEAR', 'VALUE', type = 'b') 绘制一些时间序列图. 由于缺少时间序列中年份之间的一些数据，我想编写包含年份之间的 NA 值的新列，以便我可以绘制数据而无需在 YEAR 间隙中连接线（在我的示例中：填写 NA 值1984 年到 1987 年在 BATLEY 项目中）。

有没有办法做到这一点？任何帮助将不胜感激！谢谢！

我的 .csv 文件如下所示：

NAME;                ID;      YEAR; VALUE
NAME1;              885;    1988;      -2
NAME1;              885;    1989;       0
NAME2;  2665;   1999;       4
NAME2;  2665;   2000;       8
NAME2;  2665;   2001;      19
NAME2;  2665;   2002;      13
NAME2;  2665;   2003;      13
NAME3;          893 ;   1983;       0
NAME3;          893 ;   1988;       2
NAME3;          893 ;   1989;      -1
NAME4;              877 ;   1972;      -1
NAME5;          894 ;   1973;      -3

Answer 1

您可以使用 sep=";" 来读取您所显示的文件在 read.csv 中识别单独的值。 您可能会考虑使用类似下面的代码来读取数据、修复日期、添加 NA 并绘制绘图。我将您的数据放在一个名为“plot_test.txt”的文件中，因此 read.csv 从那里获取数据。另外，根据您对 BATLEY 项目的评论，我假设您想为情节中的每个项目分开行。

# read data file into xx and change character years to Date values
xx <- read.csv("plot_test.txt",header=TRUE,sep=";")
xx$YEAR <- as.Date(paste(as.character(xx$YEAR),"-01-01",sep=""))

#    create df as a template for all years and names
date_seq <- seq(min(xx$YEAR),max(xx$YEAR),by="12 month")
df <- merge(data.frame(NAME=unique(xx$NAME),ID=unique(xx$ID)),data.frame(YEAR=date_seq,VALUE=NA),all=TRUE)

# create unique names in xx and df to merge on
xx$NAME_YR <- paste(xx$NAME,xx$YEAR,sep="")
df$NAME_YR <- paste(df$NAME,df$YEAR,sep="")
#  merge keeping only real data columns and restore original column names
xy <- merge(xx, df,by="NAME_YR",all=TRUE)[,c("NAME.y","ID.y","YEAR.y","VALUE.x")]
names(xy) <- names(xx)[1:4]

# plot xy using ggplot
library(ggplot2)
sp <- ggplot(data=xy, aes(x=YEAR, y=VALUE, colour=NAME)) + geom_point() + geom_line()
plot(sp)  

Answer 2

很高兴听到你想通了。我仍然想知道每页的情节数量。我在代码中添加了几行，以允许您设置显示在一页上的绘图的行数和列数，然后根据需要循环遍历尽可能多的绘图页。我还添加了一些 ggplot 东西来改变绘图文本的外观。

    # read data file into xx and change character years to Date values
    xx <- read.csv("plot_test.txt",header=TRUE,sep=";")
    xx$YEAR <- as.Date(paste(as.character(xx$YEAR),"-01-01",sep=""))
    xx$NAME_YR <- paste(xx$NAME,xx$YEAR,sep="")

    #  create Year template for years between min and max years for each NAME
    xxmin <- as.Date(tapply(xx$YEAR, xx$NAME, min ), origin="1970-01-01")
    xxmax <- as.Date(tapply(xx$YEAR, xx$NAME, max ), origin="1970-01-01")
    xxdates <- mapply(seq, xxmin, xxmax, by="12 month")
    xxyears <- data.frame(NAME=rep(names(xxdates), sapply(xxdates, length)), 
                          YEAR=as.Date(unlist (xxdates),origin="1970-01-01"))
    xxyears$NAME_YR <- paste(xxyears$NAME,xxyears$YEAR,sep="")

    #  merge  template and data and assign colnames to ploting data
    xy <- merge(xx, xxyears, by="NAME_YR", all=TRUE)[,c("NAME.y","ID","YEAR.y","VALUE")]
    names(xy) <- c("NAME","ID","YEAR","VALUE")

   # plot each NAME in a separate chart with own time axis
   library(ggplot2)
   rows_pg <- 2   # number of rows of plots per page
   cols_pg <- 2   # number of columns of plots per page
   chts_pg <- rows_pg*cols_pg
   num_plots <- nlevels(xy$NAME)
   #  set plot axis labels and main titles and set values for text
   spttl <- ggtitle("Your plot title\nSecond line of your plot title")
   spaxlb <- labs ( x="Year", y="Data Values")  
   spth <-  theme(plot.title=element_text(size=16, face="bold", colour="blue") )
   spth <-  spth + theme(axis.title.x= element_text(size=14, colour="blue") )
   spth <-  spth + theme(axis.title.y = element_text(size=14, colour="blue") )
   spth <-  spth + theme(axis.text = element_text(size=14, colour="black") )
   spth <-  spth + theme(strip.text = element_text(size=14, colour="brown"))
   #  generate plots
   for( iplt in seq(1,num_plots, chts_pg) )  {
     sp <- ggplot(data=xy[xy$NAME %in% levels(xy$NAME)[iplt:(iplt+chts_pg-1)], ],
                  aes(x=YEAR, y=VALUE)) + geom_line() + geom_point()
     sp  <- sp + facet_wrap(~ NAME, scales="free_x", nrow=rows_pg, ncol=cols_pg)
     plot(sp + spttl + spaxlb + spth)
   }