Ok, Sorry for botched up prior post. Fixed data on site re 9/10/04 and
later ES SP mini data and fixed date problem. Only one new observation
on 10/1 since last time. Current situation open. Here is result:
DATE Y M D W SP VIX SP10AHD
2185 09/03/04 2004 9 3 5 1114.40 13.91 7.85
2189 09/10/04 2004 9 10 5 1123.40 13.76 -11.90
2190 09/13/04 2004 9 13 1 1128.25 13.17 -23.00
2204 10/01/04 2004 10 1 5 1133.25 12.75 -25.00
2227 11/03/04 2004 11 3 3 1145.00 14.04 NA
2228 11/04/04 2004 11 4 4 1157.00 13.45 NA
> cat('NumObs is '); nrow(OBSALL); summary(OBSALL$SP10AHD)
NumObs is [1] 288
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
-97.9000 -21.1300 -0.7000 -0.9738 22.6000 77.0000 21.0000
> OBSNEW<- SOGI[MINNEW,]; tail(OBSNEW);
DATE Y M D W SP VIX SP10AHD
2005 12/16/03 2003 12 16 2 1071.6 15.93 37.4
2052 02/25/04 2004 2 25 3 1141.4 14.93 -22.7
2079 04/02/04 2004 4 2 5 1141.9 15.64 -8.4
2117 05/27/04 2004 5 27 4 1122.5 15.28 2.9
2178 08/25/04 2004 8 25 3 1103.8 14.98 13.8
2227 11/03/04 2004 11 3 3 1145.0 14.04 NA
#Gregmisc is updated in 2.0 and instead there appears to be gtools,
gdata, and gmodels
#didn't know which ones so I used them all.
require(gtools) require(gdata) require(gmodels) info<- function(O) list(head(O),tail(O)) lead<- function(O,N) c(diff(O,N),rep(NA,N)) SOGI<- read.csv("http://home.hawaii.rr.com/zephyr/vix.csv",header=T) SOGI<- transform(SOGI,SP10AHD=lead(SP,10)) # add SP 10 dy ahead MINALL<- SOGI$VIX==running(SOGI$VIX,,min,20,pad=T) # all VIX mins MINNEW<- MINALL&(running(MINALL,,sum,20,pad=T)==1) # fresh VIX mins OBSALL<- SOGI[MINALL,]; tail(OBSALL) cat('NumObs is '); nrow(OBSALL); summary(OBSALL$SP10AHD) OBSNEW<- SOGI[MINNEW,]; tail(OBSNEW); cat('NumObs is '); nrow(OBSNEW); summary(OBSNEW$SP10AHD).
