Converting zoo object into a weekly time series

问题

I am working on building a time series in R programming language. I m having a zoo object which is follows:

I 'd like to convert this into a weekly time series data for analysis and typed in the following code

tt2<-as.ts(zz,freq=365.25/7,start=decimal_date(ymd("2018-01-01")))
tt2[is.na(tt2)]<-0

However, I get the following output:

Time Series:
Start = 17538 
End = 18532 
Frequency = 0.142857142857143

While I'd like to see the output in line with something like this:

Time Series:
Start = c(2018,2) 
End = c(2020,40)
Frequency = 52

or since we can have both 53 and 52 weeks, something like:

Time Series:
Start = 1991.0848733744 
End = 2005.34360027378 
Frequency = 52.1785714285714 

I also tried to do the following ,

library(zoo)    
zz <- read.zoo(data, split = 1, index = 2,FUN=as.week")
   

and converted the following into the format:

However, if i try to convert this into a time series, I receive the following output:

Time Series:
Start = 2505 
End = 2647 
Frequency = 1 
  [1] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA
 [40] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA
 [79] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA
[118] NA NA NA NA NA NA NA NA NA 64 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

I'd be keen to receive your thoughts on this

回答1:

I suppose using tsibble would more easier to convert your series from daily frequency to weekly frequency. At the end you can change to zoo object again.

Here is a short code on what I done

data

# A tibble: 14 x 2
   Date            Y
   <date>      <dbl>
 1 2020-01-01 0.176 
 2 2020-01-02 0.521 
 3 2020-01-03 0.348 
 4 2020-01-04 0.801 
 5 2020-01-05 0.963 
 6 2020-01-06 0.0723
 7 2020-01-07 0.638 
 8 2020-01-08 0.842 
 9 2020-01-09 0.298 
10 2020-01-10 0.902 
11 2020-01-11 0.943 
12 2020-01-12 0.884 
13 2020-01-13 0.266 
14 2020-01-14 0.789 

   library(tsibble)
    library(tidyverse)
    library(zoo)
    data$Date<-as.Date(data$Date)
    data.w<-data%>%as_tsibble(index=Date)%>% index_by(year_week = ~ yearweek(.)) %>% summarise(weekly = sum(Y, na.rm = TRUE))
    data.z<-zoo(data.w)

> data.z
  year_week weekly  
1 2020 W01  2.809756
2 2020 W02  4.579329
3 2020 W03  1.055690

来源：https://stackoverflow.com/questions/64960531/converting-zoo-object-into-a-weekly-time-series