Search code examples
rfixedrolling-computation

Construct a "simple" fixed rolling window

Is it possible to construct a "simple" fixed rolling window? Say I have the following dataset:

         Apple Microsoft     Tesla    Amazon
 2010 0.8533719 0.8078440 0.2620114 0.1869552
 2011 0.7462573 0.5127501 0.5452448 0.1369686
 2012 0.7580671 0.5062639 0.7847919 0.8362821
 2013 0.3154078 0.6960258 0.7303597 0.6057027
 2014 0.4741735 0.3906580 0.4515726 0.1396147
 2015 0.4230036 0.4728911 0.1262413 0.7495193
 2016 0.2396552 0.5001825 0.6732861 0.8535837
 2017 0.2007575 0.8875209 0.5086837 0.2211072
#I want to be able to produce the following result
s.matrix <- x[1:4,] 
#For the next period, I want to drop the first period and add the next period: 
s.matrix <- x[2:5,] 
#For the rest of the dataset it should be:
 x[3:6,], x[4:7,], x[5:8,]
#That is, the width should always be equal to four.

I know that lapply is able to do something similar, but then I have to set a fixed value such that it only adds the new variables to an already existing matrix without removing the first observation....or am I wrong?

Solution

  • Assuming x is a data.frame as in the Note at the end, use rollapply to get the desired indexes and apply to generate the corresponding list of data frames.

    library(zoo)

apply(rollapply(1:nrow(x), 4, c), 1, function(ix) x[ix, ])

    giving:

    [[1]]
       Apple Microsoft   Tesla  Amazon
2010 0.85337   0.80784 0.26201 0.18696
2011 0.74626   0.51275 0.54524 0.13697
2012 0.75807   0.50626 0.78479 0.83628
2013 0.31541   0.69603 0.73036 0.60570

[[2]]
       Apple Microsoft   Tesla  Amazon
2011 0.74626   0.51275 0.54524 0.13697
2012 0.75807   0.50626 0.78479 0.83628
2013 0.31541   0.69603 0.73036 0.60570
2014 0.47417   0.39066 0.45157 0.13961

[[3]]
       Apple Microsoft   Tesla  Amazon
2012 0.75807   0.50626 0.78479 0.83628
2013 0.31541   0.69603 0.73036 0.60570
2014 0.47417   0.39066 0.45157 0.13961
2015 0.42300   0.47289 0.12624 0.74952

[[4]]
       Apple Microsoft   Tesla  Amazon
2013 0.31541   0.69603 0.73036 0.60570
2014 0.47417   0.39066 0.45157 0.13961
2015 0.42300   0.47289 0.12624 0.74952
2016 0.23966   0.50018 0.67329 0.85358

[[5]]
       Apple Microsoft   Tesla  Amazon
2014 0.47417   0.39066 0.45157 0.13961
2015 0.42300   0.47289 0.12624 0.74952
2016 0.23966   0.50018 0.67329 0.85358
2017 0.20076   0.88752 0.50868 0.22111

    Note

    We used this for x:

    Lines <- "         Apple Microsoft     Tesla    Amazon
 2010 0.8533719 0.8078440 0.2620114 0.1869552
 2011 0.7462573 0.5127501 0.5452448 0.1369686
 2012 0.7580671 0.5062639 0.7847919 0.8362821
 2013 0.3154078 0.6960258 0.7303597 0.6057027
 2014 0.4741735 0.3906580 0.4515726 0.1396147
 2015 0.4230036 0.4728911 0.1262413 0.7495193
 2016 0.2396552 0.5001825 0.6732861 0.8535837
 2017 0.2007575 0.8875209 0.5086837 0.2211072"

x <- read.table(text = Lines)