Different output from code used in function and used alone in R -

i've been working kind of data evaluate stocks diploma thesis

library(quantmod) getsymbols("nok",from="2012-01-01") data<-nok p1<-4  dc<-data[,4] do<-data[,1] emac<-ema(dc,n=p1) w1<-1.1 profit_l_1<-((lag(dc,-1)-lag(do,-1))/(lag(do,-1)))*100 profit_l_2<-((lag(dc,-2)-lag(do,-1))/(lag(do,-1)))*100 profit_l_3<-((lag(dc,-3)-lag(do,-1))/(lag(do,-1)))*100 profit_l_4<-((lag(dc,-4)-lag(do,-1))/(lag(do,-1)))*100 profit_l_5<-((lag(dc,-5)-lag(do,-1))/(lag(do,-1)))*100 profit_l_all<-ifelse(profit_l_1>w1,profit_l_1,                      ifelse(profit_l_2>w1,profit_l_2,                             ifelse(profit_l_3>w1,profit_l_3,                                    ifelse(profit_l_4>w1,profit_l_4,                                           ifelse(profit_l_5>w1,profit_l_5,profit_l_5))))) choosedn<-ifelse(profit_l_all==profit_l_1,1,                  ifelse(profit_l_all==profit_l_2,2,                         ifelse(profit_l_all==profit_l_3,3,                                ifelse(profit_l_all==profit_l_4,4,                                       ifelse(profit_l_all==profit_l_5,5,0)))))  profit_day<-profit_l_all/choosedn winloss<-ifelse(profit_day>0.3,1,                 ifelse(profit_day<=0.3,0,"nothing")) pos_emac_dc_d<-emac<dc pos_emac_dc_up<-emac>dc frame<-data.frame(pos_emac_dc_up,pos_emac_dc_d,profit_l_all,choosedn,profit_day,winloss) colnames(frame)<-c("pos_emac_dc_up","pos_emac_dc_d","profit_l_all","choosedn","profit_day","winloss") frame<-frame[complete.cases(frame),] 

output of data.frame called frame work with. clear without na's.

i have written function quick evaluate combinations of variables in frame

quickmycomb<-function (x,y){   combination<-deparse(substitute(y))   df<-subset(x,y)   success<-length(subset(df,winloss==1)$winloss)/length(df$winloss)   absnumber<-nrow(df)   relnumber<-nrow(df)/nrow(frame)   sl<-data.frame(success,absnumber,relnumber)   return(sl) }  

and when use function like

quickmycomb(frame,lag(pos_emac_dc_d,1) & pos_emac_dc_up) 

it says output frame 86 rows in case

    success absnumber relnumber 1 0.6162791        86 0.1396104 

but when use code function subset(frame,"combination") same combination of variables like

subset(frame,lag(pos_emac_dc_d,1) & pos_emac_dc_up) 

it says output data.frame 0 rows!!! don't understand it, suppose right output second, don't know why doesn't work in function. advices should problem?

you cannot use subset variable condition because subset use substitute see passed function. however, can pass through values subset. 1 fix is

quickmycomb.1 <- function (x,...) {   df <- subset(x,...)   success <- length(subset(df,winloss==1)$winloss)/length(df$winloss)   absnumber <- nrow(df)   relnumber <- nrow(df)/nrow(frame)   sl <- data.frame(success,absnumber,relnumber)   return(sl) } 

alternative can process subsetting inside function with

quickmycomb.2 <- function (x,y){   combination <- eval(substitute(y), x, parent.frame())   df <- x[combination, ]   success <- length(subset(df,winloss==1)$winloss)/length(df$winloss)   absnumber <- nrow(df)   relnumber <- nrow(df)/nrow(frame)   sl <- data.frame(success,absnumber,relnumber)   return(sl) }  

when run them both

quickmycomb.1(frame,lag(pos_emac_dc_d,1) & pos_emac_dc_up) quickmycomb.2(frame,lag(pos_emac_dc_d,1) & pos_emac_dc_up) 

they both produce

  success absnumber relnumber 1     nan         0         0 

just when run current environment.