For the demonstration, we'll be importing data from Yahoo with the TradingInstrument package.
julia> using TradingInstrument
julia> using TimeSeries
julia> AAPL = yahoo("AAPL");
julia> head(AAPL, 3)
3x7 DataFrame:
Date Open High Low Close Vol Adj
[1,] 2010-01-12 209.19 209.77 206.42 207.72 21230700 204.7
[2,] 2010-01-13 207.87 210.93 204.1 210.65 21639000 207.59
[3,] 2010-01-14 210.11 210.46 209.02 209.43 15460500 206.38When dealing with time series data, it is often useful to lag or lead data. This
is done with the lag and lead functions. The lag function allows negative
integers, in which case it simply calls the lead function.
julia> head(lag!(AAPL,"Close"), 3)
3x8 DataFrame:
Date Open High Low Close Vol Adj Close_lag_1
[1,] 2010-01-12 209.19 209.77 206.42 207.72 21230700 204.7 NA
[2,] 2010-01-13 207.87 210.93 204.1 210.65 21639000 207.59 207.72
[3,] 2010-01-14 210.11 210.46 209.02 209.43 15460500 206.38 210.65
Other important functions include the moving function, which calculates data within
a rolling window of consecutive rows of data, upto, which calculates all the data from
the beginning to the current row, and two returns functions that calculate the percent
change in a value between rows (simple and logarithmic are available).
The indexdate.jl file includes functions that aggregate a DataFrame whose first row is a Calendar
type and does so along a specified time period. indexmonth for example will aggregate rows whose month
value is whatever is specified in the function call.
If you're interested in running the test suite, you can call the @timeseries macro inside a Julia session.
julia> @timeseries
Running tests:
** test/returns.jl
** test/lag.jl
** test/moving.jl
** test/upto.jl
** test/indexdate.jl