Advance a Time Series
Create a new series with all values advanced forward one period.
The value of period 1, becomes the value at period 2,
value at 2 becomes the original value at 3, etc. The opposite
of Lag. NA is used to fill.
Next(x, k = 1) ## S3 method for class 'quantmod.OHLC' Next(x,k=1) ## S3 method for class 'zoo' Next(x,k=1) ## S3 method for class 'data.frame' Next(x,k=1) ## S3 method for class 'numeric' Next(x,k=1)
x |
vector or series to be advanced |
k |
periods to advance |
Shift series k-periods up, appending NAs to end of series.
Specifically designed to handle quantmod.OHLC and
zoo series within the quantmod workflow.
If no S3 method is found, a call to lag in base is made,
with the indexing reversed to shift the time series forward.
The original x appended with k NAs and
missing the leading k values.
The returned series maintains the number of obs. of the original.
Unlike Lag, only one value for k is allowed.
This function's purpose is to get the “next” value of
the data you hope to forecast, e.g. a stock's closing value
at t+1. Specifically to be used within
the quantmod framework of specifyModel, as a
functional wrapper to the LHS of the model equation.
It is not magic - and thus will not get tomorrow's values...
Jeffrey A. Ryan
Stock.Close <- c(102.12,102.62,100.12,103.00,103.87,103.12,105.12) Close.Dates <- as.Date(c(10660,10661,10662,10665,10666,10667,10668),origin="1970-01-01") Stock.Close <- zoo(Stock.Close,Close.Dates) Next(Stock.Close) #one period ahead Next(Stock.Close,k=1) #same merge(Next(Stock.Close),Stock.Close) ## Not run: # a simple way to build a model of next days # IBM close, given todays. Technically both # methods are equal, though the former is seen # as more intuitive...ymmv specifyModel(Next(Cl(IBM)) ~ Cl(IBM)) specifyModel(Cl(IBM) ~ Lag(Cl(IBM))) ## End(Not run)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.