rcsvvectorirrxirr

How to run a function on multiple csv's at the same time

I have a set of 3 csv's which are all saved in the below directory:

setwd("~/R/CashFlows")

The csv's are named:

"Cashflows1.csv" "Cashflows2.csv" "Cashflows3.csv"

And are all in a form similar to the below (the below example is "Cashflows1.csv")

19/10/2003  -13275
19/11/2003  940.49
19/12/2003  884.71
19/01/2004  832.11
19/02/2004  782.49
19/03/2004  735.74
19/04/2004  691.64
19/05/2004  650.09
19/06/2004  610.91
19/07/2004  573.99
19/08/2004  539.2
19/09/2004  506.42
19/10/2004  475.54
19/11/2004  441.05
19/12/2004  413.91
19/01/2005  388.37
19/02/2005  364.31
19/03/2005  341.66
19/04/2005  320.34
19/05/2005  300.28
19/06/2005  281.39
19/07/2005  263.63
19/08/2005  246.91
19/09/2005  231.2
19/10/2005  216.41
19/11/2005  202.51
19/12/2005  189.43
19/01/2006  177.15
19/02/2006  165.6
19/03/2006  154.75
19/04/2006  144.55
19/05/2006  134.98
19/06/2006  125.99
19/07/2006  117.55
19/08/2006  109.62
19/09/2006  102.18

Or in Vector form:

dat <- read.csv("cashflows1.csv", skip=1, header=F)$V2
> dat
 [1] -13275.00    940.49    884.71    832.11    782.49    735.74    691.64    650.09    610.91    573.99    539.20
[12]    506.42    475.54    441.05    413.91    388.37    364.31    341.66    320.34    300.28    281.39    263.63
[23]    246.91    231.20    216.41    202.51    189.43    177.15    165.60    154.75    144.55    134.98    125.99
[34]    117.55    109.62    102.18

I have created the following function to return the annualized IRR of each cashflow file (Cashflows1.csv is used in the example)

 setwd("~/R Studio/Technical Test")
> dat <- read.csv("cashflows1.csv", skip=1, header=F)$V2
> npv<-function(i,cf,t=seq(along=cf)) sum (cf/(1+i)^t)
> irr <- function(cf) {uniroot(npv, c(0,1), cf=cf)$root }
> irr(dat)
[1] 0.002384391
> var <- irr(dat)
> 
> AIRR <- (1+var)^12-1
> AIRR
[1] 0.02899093
>

How would i go about changing the function such that it calculates the IRR of all the csv's in the directory simultaneously as i currently have to stipulate the specific cashflow file for the "dat" variable. But would like it to calculate it for an "X" number of csv's.

Can anyone help me?

Solution

  • We have multiple things to tackle here but all are solvable with similar tools, the apply-family of functions.

    Let's go at it in steps:

    1. You need to read in multiple csv-files at once, an answer to this can be easily found here:

    How to import multiple .csv files at once?

    from this we can take @A5C1D2H2I1M1N2O1R2T1 excellent code to create a list of data frames with all csvs accounted for:

    temp = list.files(pattern="*.csv")
myfiles = lapply(temp, read.delim)

    Please note the conditions and comments to this functions in the original post!

    1. Next we want to again "apply" your defined function on all data frames in the list. So the solution can be found here:

    Same function over multiple data frames in R

    Using lapply we can take in a list of data frames (defined in step 1) and apply one function to each data frame in the list like this:

    result <- lapply(myfiles, function(x) {
          uniroot(npv, c(0,1), x=x)$root
})

    Note that you still need to define the function npvahead of time but not irr, the result will be a list of the outputs.