How to calculate percentage change from different rows over different spans

Question

I am trying to calculate the percentage change in price for quarterly data of companies recognized by a gvkey(1001, 1384, etc...). and it\'s corresponding quart

Answer 1

df <- read.table(text="gvkey  PRCCQ
1  1004  5.500
2  1004  4.450
3  1004  4.500
4  1004  8.010
5  1005  4.450
6  1005  4.500",header=TRUE)

library(plyr)
library(quantmod)
ddply(df, "gvkey", transform, DeltaCol = Delt(PRCCQ,k=3))
#error

Delt2 <- function(x,k) {
  if(length(x)>k) as.vector(Delt(x1=x,k=k)) else rep(NA,length(x))
}

ddply(df, "gvkey", transform, DeltaCol = Delt2(PRCCQ,k=3))
#  gvkey PRCCQ  DeltaCol
#1  1004  5.50        NA
#2  1004  4.45        NA
#3  1004  4.50        NA
#4  1004  8.01 0.4563636
#5  1005  4.45        NA
#6  1005  4.50        NA

Answer 2

You can declare your data as ts() and use cbind() and diff()

data <- read.table(header=T,text='gvkey  PRCCQ
   1004 23.750
   1004 13.875
   1004 11.250
   1004 10.375
   1004 13.600
   1004 14.000
   1004 17.060
   1005  8.150
   1005  7.400
  1005 11.440
  1005  6.200
  1005  5.500
  1005  4.450
  1005  4.500
  1005  8.010')

data <- split(data,list(data$gvkey))
(newdata <- do.call(rbind,lapply(data,function(data) { data <- ts(data) ; cbind(data,Quarter=diff(data[,2]),Two.Quarter=diff(data[,2],2))})))

      data.gvkey data.PRCCQ Quarter Two.Quarter
 [1,]       1004     23.750      NA          NA
 [2,]       1004     13.875  -9.875          NA
 [3,]       1004     11.250  -2.625     -12.500
 [4,]       1004     10.375  -0.875      -3.500
 [5,]       1004     13.600   3.225       2.350
 [6,]       1004     14.000   0.400       3.625
 [7,]       1004     17.060   3.060       3.460
 [8,]       1005      8.150      NA          NA
 [9,]       1005      7.400  -0.750          NA
[10,]       1005     11.440   4.040       3.290
[11,]       1005      6.200  -5.240      -1.200
[12,]       1005      5.500  -0.700      -5.940
[13,]       1005      4.450  -1.050      -1.750
[14,]       1005      4.500   0.050      -1.000
[15,]       1005      8.010   3.510       3.560

EDIT:

Another way, without split() and lapply() (probably faster)

data <- read.table(header=T,text='gvkey  PRCCQ
       1004 23.750
       1004 13.875
       1004 11.250
       1004 10.375
       1004 13.600
       1004 14.000
       1004 17.060
       1005  8.150
       1005  7.400
      1005 11.440
      1005  6.200
      1005  5.500
      1005  4.450
      1005  4.500
      1005  8.010')
newdata <- do.call(rbind,by(data, data$gvkey,function(data) { data <- ts(data) ; cbind(data,Quarter=diff(data[,2]),Two.Quarter=diff(data[,2],2))}))