rivercode
2011-Jan-10 00:10 UTC
[R] Aggragating subsets of data in larger vector with sapply
Have 40,000 rows of buy/sell trade data and am trying to add up the buys for each second, the code works but it is very slow. Any suggestions how to improve the sapply function ? secEP = endpoints(xSym$Direction, "secs") # vector of last second on an XTS timeseries object with multiple entries for each second. d = xSym$Direction s = xSym$Size buySize = sapply(1:(length(secEP)-1), function(y) { i = (secEP[y]+ 1):secEP[y+1]; # index of vectors between each secEP return(sum(as.numeric(s[i][d[i] == "buy"]))); } ) Object details: secEP = numeric Vector of one second Endpoints in xSym$Direction.> head(xSym$Direction)Direction 2011-01-05 09:30:00 "unkn" 2011-01-05 09:30:02 "sell" 2011-01-05 09:30:02 "buy" 2011-01-05 09:30:04 "buy" 2011-01-05 09:30:04 "buy" 2011-01-05 09:30:04 "buy"> head(xSym$Size)Size 2011-01-05 09:30:00 " 865" 2011-01-05 09:30:02 " 100" 2011-01-05 09:30:02 " 100" 2011-01-05 09:30:04 " 100" 2011-01-05 09:30:04 " 100" 2011-01-05 09:30:04 " 41" Thanks, Chris -- View this message in context: http://r.789695.n4.nabble.com/Aggragating-subsets-of-data-in-larger-vector-with-sapply-tp3206445p3206445.html Sent from the R help mailing list archive at Nabble.com.
Jim Holtman
2011-Jan-10 01:50 UTC
[R] Aggragating subsets of data in larger vector with sapply
split the data by truncating the time to a second, then process each group. this will save the subsetting you are doing. also merge the data with direction and size in the same frame. it looks like you can subset by "buy" to begin with. Sent from my iPad On Jan 9, 2011, at 19:10, rivercode <aquanyc at gmail.com> wrote:> > > Have 40,000 rows of buy/sell trade data and am trying to add up the buys for > each second, the code works but it is very slow. Any suggestions how to > improve the sapply function ? > > secEP = endpoints(xSym$Direction, "secs") # vector of last second on an XTS > timeseries object with multiple entries for each second. > d = xSym$Direction > s = xSym$Size > buySize = sapply(1:(length(secEP)-1), function(y) { > i = (secEP[y]+ 1):secEP[y+1]; # index of vectors between each secEP > return(sum(as.numeric(s[i][d[i] == "buy"]))); > } ) > > Object details: > > secEP = numeric Vector of one second Endpoints in xSym$Direction. > >> head(xSym$Direction) > Direction > 2011-01-05 09:30:00 "unkn" > 2011-01-05 09:30:02 "sell" > 2011-01-05 09:30:02 "buy" > 2011-01-05 09:30:04 "buy" > 2011-01-05 09:30:04 "buy" > 2011-01-05 09:30:04 "buy" > >> head(xSym$Size) > Size > 2011-01-05 09:30:00 " 865" > 2011-01-05 09:30:02 " 100" > 2011-01-05 09:30:02 " 100" > 2011-01-05 09:30:04 " 100" > 2011-01-05 09:30:04 " 100" > 2011-01-05 09:30:04 " 41" > > Thanks, > Chris > > > -- > View this message in context: http://r.789695.n4.nabble.com/Aggragating-subsets-of-data-in-larger-vector-with-sapply-tp3206445p3206445.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code.
Joshua Ulrich
2011-Jan-12 04:35 UTC
[R] Aggragating subsets of data in larger vector with sapply
Hi Chris, This seems to work on the sample data you provided. FUN <- function(x) { x <- xts(as.numeric(x),index(x)) period.apply(x, endpoints(x,"secs"), sum) } lapply(split.default(xSym$Size,xSym$Direction), FUN) Best, -- Joshua Ulrich ?| ?FOSS Trading: www.fosstrading.com On Sun, Jan 9, 2011 at 6:10 PM, rivercode <aquanyc at gmail.com> wrote:> > > Have 40,000 rows of buy/sell trade data and am trying to add up the buys for > each second, the code works but it is very slow. ?Any suggestions how to > improve the sapply function ? > > secEP = endpoints(xSym$Direction, "secs") ?# vector of last second on an XTS > timeseries object with multiple entries for each second. > d = xSym$Direction > s = xSym$Size > buySize = sapply(1:(length(secEP)-1), function(y) { > ? ? ? ?i = ?(secEP[y]+ 1):secEP[y+1]; # index of vectors between each secEP > ? ? ? ?return(sum(as.numeric(s[i][d[i] == "buy"]))); > } ) > > Object details: > > secEP = numeric Vector of one second Endpoints in xSym$Direction. > >> head(xSym$Direction) > ? ? ? ? ? ? ? ? ? ?Direction > 2011-01-05 09:30:00 "unkn" > 2011-01-05 09:30:02 "sell" > 2011-01-05 09:30:02 "buy" > 2011-01-05 09:30:04 "buy" > 2011-01-05 09:30:04 "buy" > 2011-01-05 09:30:04 "buy" > >> head(xSym$Size) > ? ? ? ? ? ? ? ? ? ?Size > 2011-01-05 09:30:00 " 865" > 2011-01-05 09:30:02 " 100" > 2011-01-05 09:30:02 " 100" > 2011-01-05 09:30:04 " 100" > 2011-01-05 09:30:04 " 100" > 2011-01-05 09:30:04 " ?41" > > Thanks, > Chris > > > -- > View this message in context: http://r.789695.n4.nabble.com/Aggragating-subsets-of-data-in-larger-vector-with-sapply-tp3206445p3206445.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >