将多个不规则时间序列转换为规则时间序列 [英] Convert multiple irregular time series into regular time series

问题描述

我有一个由多个不规则时间序列(data.frame)组成的 data.frame，看起来像这样

I have a data.frame of multiple irregular time series (data.frame) which looks like this

station   Time     WaterTemp
1       01-01-1974  5.0000000
1       01-02-1974  5.0000000
1       01-03-1974  8.6000004
1       01-05-1974  8.1333332
1       01-07-1974  12.7999999
2       01-01-1974  5.0000000
2       01-02-1974  5.0000000
2       01-04-1974  8.6000004
2       01-06-1974  8.1333332
2       01-08-1974  12.7999999

我想将其转换为常规时间序列 (ts) 对象，它应该如下所示

I want to convert this into regular time series (ts) object which should look like this

Time        Staion1     Station2
 Jan1974    5.0000000  5.0000000
 Feb1974    5.0000000  5.0000000
 Mar1974    8.6000004  NA
 Apr1974    NA         8.6000004
 May1974    8.1333332  NA
 June1974   NA         8.1333332
 July1974   12.7999999 NA
  Aug1974  NA         12.7999999
  Sep1974  NA         NA
  Oct1974  7.9         NA
  Nov1974  NA         NA
  Dec1974  NA         7.4

我该怎么做?虽然针对单个时间序列有很多解决方案，但我还没有遇到过处理多个时间序列的方法.

How do I do that? Although there are lots of solutions for a single time series, but I haven't come across one dealing with multiple time series.

谢谢，

推荐答案

如果 DF 是你的数据框，那么试试这个.在最后一行转换为 ts 使其成为常规，然后我们转换回动物园:

If DF is your data frame then try this. Converting to ts in the last line makes it regular and then we convert back to zoo:

library(zoo)
z <- read.zoo(DF, split = 1, index = 2, format = "%d-%m-%Y")
z.ym <- aggregate(z, as.yearmon, identity) # convert to yearmon
zm <- aggregate(as.zoo(as.ts(z.ym)), as.yearmon, identity)

最后一行的替代方法是这两行:

An alternative to the last line would be these two lines:

g <- zoo(, seq(start(z.ym), end(z.ym), deltat(z.ym))) # grid
zm <- merge(z.ym, g)

无论哪种情况，此时 coredata(zm) 是数据部分，time(zm) 是索引，尽管您可能希望将其保留为动物园对象，以便您可以使用它的其他时间序列工具和许多其他接受该形式时间序列的包.

In either case, at this point coredata(zm) is the data part and time(zm) is the index although you might want to keep it as a zoo object so that you can use its other time series facilities and the many other packages which accept time series of that form.

注意:这是一个完整的独立可复制示例:

Note: Here is a complete self-contained reproducible example:

DF <- structure(list(station = c(1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 
2L), Time = structure(c(1L, 2L, 3L, 5L, 7L, 1L, 2L, 4L, 6L, 8L
), .Label = c("01-01-1974", "01-02-1974", "01-03-1974", "01-04-1974", 
"01-05-1974", "01-06-1974", "01-07-1974", "01-08-1974"), class = "factor"), 
    WaterTemp = c(5, 5, 8.6000004, 8.1333332, 12.7999999, 5, 
    5, 8.6000004, 8.1333332, 12.7999999)), .Names = c("station", 
"Time", "WaterTemp"), class = "data.frame", row.names = c(NA, 
-10L))

library(zoo)
z <- read.zoo(DF, split = 1, index = 2, format = "%d-%m-%Y")
z.ym <- aggregate(z, as.yearmon, identity) # convert to yearmon
zm <- aggregate(as.zoo(as.ts(z.ym)), as.yearmon, identity)

给予:

> zm
                 1         2
Jan 1974  5.000000  5.000000
Feb 1974  5.000000  5.000000
Mar 1974  8.600000        NA
Apr 1974        NA  8.600000
May 1974  8.133333        NA
Jun 1974        NA  8.133333
Jul 1974 12.800000        NA
Aug 1974        NA 12.800000

更新一些更正和改进.

这篇关于将多个不规则时间序列转换为规则时间序列的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！