R xts Time series, subset by last workday of the week

Question

I have daily closing price in a xts object. What I am trying to do is get the last closing price of the week. Most of the time this will be Friday but sometimes this might be Thursday or even more rare Wednesday.

I been trying to subset this with the solution found here but with no luck. Has anyone seen or found a solution for something around this.

Thanks in advance.

Please provide a reproducible example. – LyzandeR Sep 04 '15 at 22:16 — LyzandeR, Sep 04 '15 at 22:16

score 2 · Accepted Answer · edited Sep 05 '15 at 15:41

2

Convert the time series to weekly OHLC with to.weekly. The weekly closing price should be from the last trading day in the week -- be it Thursday or Wednesday.

For example, 7/4/2014 (a US holiday) was on a Friday, so there was no trading that day. The following code returns observations for 6/27 (a Friday), 7/3 (the Thursday close), and 7/11 (a Friday).

getSymbols("SPY")
Cl(to.weekly(SPY))['2014-06-27/2014-07-11']
#            SPY.Close
# 2014-06-27    195.82
# 2014-07-03    198.20
# 2014-07-11    196.61

edited Sep 05 '15 at 15:41

Joshua Ulrich

173,410
32
338
418

answered Sep 04 '15 at 23:55

Matt Brigida

191
7

Awesome!! I love this site! Thanks for your help – Aaron Soderstrom Sep 05 '15 at 15:56
1

@AaronSoderstrom Great! Feel free to accept either answer if they are both working for you. – Matt Brigida Sep 05 '15 at 20:55

score 1 · Answer 2 · answered Sep 04 '15 at 23:52

If your data only contains weekdays (no weekends), then you can use this:

require(quantmod)
getSymbols("SPY")
spy <- do.call(rbind, lapply(split(SPY, "weeks"), last))
table(weekdays(index(spy)))
#   Friday Thursday 
#      436       17

If your data contains weekends, then you need to remove them before doing the split, lapply, rbind.

data(sample_matrix)
x <- as.xts(sample_matrix)
x <- x[.indexwday(x) %in% 1:5]
y <- do.call(rbind, lapply(split(x, "weeks"), last))

R xts Time series, subset by last workday of the week

2 Answers2