logging - Normal and log return plotting in R and corresponding returns -

i have data sheet, lot of var. companies dax index. i'm trying time series bmw , siemens , log return them. sorry, can not attach pics... sample:

positions   allianz.hldg.   commerzbank dresdner.bank   bmw schering    basf 01/01/1973  155.51  147.41  18.40   103.97  36.88   14.96 01/02/1973  155.51  147.41  18.40   109.05  37.44   15.43 01/03/1973  160.58  149.14  18.80   109.83  37.79   15.61 01/04/1973  162.27  152.05  18.91   110.81  37.86   15.85 01/05/1973  164.30  152.05  18.89   109.44  37.44   15.75 01/08/1973  164.30  152.25  18.99   109.05  37.79   15.80 

the datas 197x, need datas 02.01.1989, till end of sheet.

the other extract data 1996 (bmw, siemens). , calculate corresponding returns on specific year. i'm totally stuck , appreciate help. thanks!

please me how calculate in r.

you can quantmod package. calling table above df,

library(quantmod) df$positions <- as.date(df$positions, format="%m/%d/%y") rownames(df) <- df$positions bmw.xts <- xts(df$bmw,order.by=df$positions) periodreturn(bmw.xts,period="daily") #            daily.returns # 1973-01-01   0.000000000 # 1973-01-02   0.048860248 # 1973-01-03   0.007152682 # 1973-01-04   0.008922881 # 1973-01-05  -0.012363505 # 1973-01-08  -0.003563596  periodreturn(bmw.xts,period="daily", type="log") #            daily.returns # 1973-01-01   0.000000000 # 1973-01-02   0.047704097 # 1973-01-03   0.007127223 # 1973-01-04   0.008883307 # 1973-01-05  -0.012440569 # 1973-01-08  -0.003569961  periodreturn(bmw.xts,period="daily", subset="1973-01-03::") #            daily.returns # 1973-01-03   0.007152682 # 1973-01-04   0.008922881 # 1973-01-05  -0.012363505 # 1973-01-08  -0.003563596 

the way table set up, first column, containing dates, loaded column positions, not want. nevertheless left way avoid confusion. positions timestamp in example. these have converted dates as.date(...) function. then, create xts time series, bmw.xts bmw column. can use quantmod function periodreturn(...) generate returns. subset= string formatted "start::end". if end missing, time series taken last row.

finally, example of plotting

library(ggplot2) library(reshape2)   # melt(...) basf.xts <- xts(df$basf,order.by=df$positions) returns <- data.frame(df$positions,                       periodreturn(bmw.xts,period="daily",type="log"),                       periodreturn(basf.xts,period="daily",type="log")) colnames(returns)=c("date","bmw","basf") gg <- melt(returns, id="date", variable.name="stock", value.name="log.return") ggplot(gg, aes(x=date, y=log.return, color=stock))+   geom_point()+ geom_line()