CODEWITHSUNDEEP
rlisttime-serieslapply
erc
erc

Reputation: 10133

Convert all columns in list of data frames to time series

I have such a list:

df1 <- data.frame(var1 = 1:5, var2 = 6:10)
rownames(df1) <- 2001:2005
df2 <- data.frame(var5 = 21:25, var6 = 26:30)
rownames(df2) <- 2006:2010
mylist <- list(df1,df2)

> mylist
[[1]]
     var1 var2
2001    1    6
2002    2    7
2003    3    8
2004    4    9
2005    5   10

[[2]]
     var5 var6
2006   21   26
2007   22   27
2008   23   28
2009   24   29
2010   25   30

How can I convert each column within each data frame to a time series where start and end are given by min and max of the rownames of the respective data frame?

I tried:

lapply(mylist, function(x) {apply(x, 2, function(y) ts(y, start = min(rownames(y), end = max(rownames(y)))))})

resulting in: 

 Error in if (nobs != ndata) data <- if (NCOL(data) == 1) { : 
  missing value where TRUE/FALSE needed

but fail to make any sense of that.

Upvotes: 1

Views: 1575

Answers (1)

akrun
akrun

Reputation: 887118

We can do this by looping columns using lapply instead of apply as the output of apply is a matrix and all the class gets lost. In addition, min/max works on numeric/integer elements, so it is advisable to convert the character class of row.names to numeric

lst1 <- lapply(mylist, function(x) lapply(x, function(y) ts(y, 
      start = min(as.numeric(row.names(x))), end = max(as.numeric(row.names(x))))))
lst1[[1]][[1]]
#Time Series:
#Start = 2001 
#End = 2005 
#Frequency = 1 
#[1] 1 2 3 4 5

If we need to create ts as columns, then assign the output back to the data.frame to keep the structure as before

 lst2 <- lapply(mylist, function(x) {
       x[] <- lapply(x, function(y) ts(y, start = min(as.numeric(row.names(x))),
                     end = max(as.numeric(row.names(x)))))
         x})
 str(lst2)
 #List of 2
 #$ :'data.frame':       5 obs. of  2 variables:
 # ..$ var1: Time-Series [1:5] from 2001 to 2005: 1 2 3 4 5
 # ..$ var2: Time-Series [1:5] from 2001 to 2005: 6 7 8 9 10
 #$ :'data.frame':       5 obs. of  2 variables:
 # ..$ var5: Time-Series [1:5] from 2006 to 2010: 21 22 23 24 25
 # ..$ var6: Time-Series [1:5] from 2006 to 2010: 26 27 28 29 30

Upvotes: 3

Related Questions