No error but not able to get desired results in R - r

Question

I have been trying to find the percentage price jump (+-15 % change) in the numbers and when there is the jump it will give me the corresponding date. However, when there is a first jump it will break-out from the inner-for loop gives me the correspndong data for it and which will get store in temp.cyc data frame.
The program is not showing any error but it is repeating the same no. throughout and when I check the value of test.df it is showing NA.
Can anyone help me understand what's going on? And I am new to R so it will helpful if you can give your answer in detail. Thank You :)
# Defining variables
row= nrow(price.close)
col=ncol(price.close)
#Defining Matrix
m<-matrix(0,ncol=1,nrow=row)
p<-matrix(0,ncol=5,nrow=row)
# Dataframe to temporaily store percentage Change
test.df<- vector(mode="numeric", length=nrow(price.close))
# Dataframe to extract required Values
temp.cyc<-as.data.frame(p)
colnames(temp.cyc)<-c("cyc.duration","Start.date","End.date","Start.date.value","End.date.value")
for( j in 1:row)
{
for(i in j:row)
{
test.df<-(price.close[(i+1),2]-price.close[j,2])/price.close[j,2]
if(test.df >= 0.15 | test.df <= -0.15 | is.na(test.df)== TRUE )
{
temp.cyc$Start.date.value = price.close[j,2]
temp.cyc$End.date.value <- price.close[i,2]
temp.cyc$Start.date <- price.close[j,1]
temp.cyc$End.date <- price.close[i,1]
}
break
}
}

Answer 1

Seems to me you are using this for financial data , e.g. stock prices. If this assumption is right then I suggest that you should use packages that have this functionality. I would suggest quantmod package.
Here's short example how to get dates when price goes more then 15% up or down.
library(quantmod)
# create some dummy stock data over 10 days period
# next time I hope you will attach some of your data
stockClose <- c(100,50,75,70,68,100,115,120,130,100)
stockDates <- seq(as.Date("2014-01-01"),length=10,by="+1 days")
stock.xts <- as.xts(stockClose,stockDates)
# calculate change , check '?Delt' help for more info
change <- Delt(stock.xts)
#get only those rows where price change in both directions is higher then 15%
specialDays <- change[coredata(change) < -0.15 | coredata(change) > 0.15,]
#get dates
justDates <- index(specialDays)
which gives us "2014-01-02" ,"2014-01-03" ,"2014-01-06" and "2014-01-10"
In case you want to actually compare all possible combinations for entering and exiting position then you can use something like this :
library(quantmod)
calculatePeriods <- function(){
stockClose <- c(100,50,75,70,68,100,115,120,130,100)
stockDates <- seq(as.Date("2014-01-01"),length=10,by="+1 days")
stock.xts <- as.xts(stockClose,stockDates)
# you will be adding rows to thid df
result <- data.frame()
for(i in 1:(length(stock.xts)-1)){
for(j in 2:length(stock.xts)){
change <- (coredata(stock.xts[j])-coredata(stock.xts[i]))/coredata(stock.xts[i])
if(change < (-0.15) | change > (0.15)){
row <- data.frame("cyc.duration"=as.numeric(index(stock.xts[j])-index(stock.xts[i]),units="days"),"Start.date"=index(stock.xts[i]),"End.date"=index(stock.xts[j]),"Start.date.value"=coredata(stock.xts[i]),"End.date.value"=coredata(stock.xts[j]))
result <- rbind(result,row)
}
}
}
return(result)
}