r - Time series forecasting of price of financial instrument -

i working financial instrument's intraday time series data. have predict price of financial instrument on basis of statistical parameters(var1, var2, var3) , of time series intraday(obeserv.1, observ.2.......observ.80) data of previous period. have predict price of financial instrument in 81st period.

all lines in table mixed information in i-line useless prediction of j-line.

i planning solve problem using r. new financial modelling field. approach can take prediction. please me out this.

data set looks that

sample data

in past i've done kind of thing living. general concept think of model have predictive power, fit e.g. first half of data set model, , test see whether model has predictive power on second half of data set.

if haven't tried before, place start arma models, (see e.g. autoregressive–moving-average model on wikipedia).