Question

我在列中添加了date，hour和minute的数据框。我想通过某种时间戳绘制value列。无论如何要做到这一点？

> head(fl)
        date hour min value
1 2014-02-23     0    0    81
2 2014-02-23     0    1    65
3 2014-02-23     0    2    73
4 2014-02-23     0    3    81
5 2014-02-23     0    4    89
6 2014-02-23     0    5    69
...

现在我正在使用ggplot2，但它将每小时和每天的分钟结合在一起：（

ggplot( fl, aes( min,  value) ) + geom_line()

有什么想法吗？

Answer 1

as.POSIXct替代方案与@RobertKrzyzanowski提供相同的结果

fl <- data.frame(date=c('2014-02-23', '2014-02-22'), hour = c(0,0), min = c(1,2))
fl$stamp <- with(fl, as.POSIXct( paste(date,hour,min), format="%Y-%m-%d %H %M")) 

#> fl
#        date hour min               stamp
#1 2014-02-23    0   1 2014-02-23 00:01:00
#2 2014-02-22    0   2 2014-02-22 00:02:00

Answer 2

library(lubridate)
fl$datetime <- ymd_hm(paste(fl$date,fl$hour,fl$min,sep='-'))
ggplot(fl, aes( datetime, value) ) + geom_line()

Answer 3

尝试功能ISOdatetime(Year, Month, Day, Hour, Min, Sec)：

fl <- data.frame(date=c('2014-02-23', '2014-02-22'), hour = c(0,0), min = c(1,2))

zip <- function(x) do.call(Map, append(list(c), x))
args <- unname(append(zip(strsplit(as.character(fl$date), '-')), list(fl$hour, fl$min, 0)))
fl$timestamp <- do.call(ISOdatetime, args)

print(fl)
#         date hour min           timestamp
# 1 2014-02-23    0   1 2014-02-23 00:01:00
# 2 2014-02-12    0   2 2014-02-12 00:02:00

Answer 4

df<- within(df[5:6], { DT=format(as.POSIXct(paste(df$date, df$time, sep = ' ')),
                                 "%Y-%m-%d %H:%M:%S") 
                     })

从dataframe列创建时间戳，并在plot中用作xaxis

4 个答案: