ANN: Timestamps

[Milktrader]

The newly registered Timestamps packages provides time-aware data structures without the constraints of regular time series data. 

These *typical* constraints include no duplicate dates and sorted rows and are implemented in TimeSeries.TimeArray. Timestamps is the Wild West of time series data where you can do more things such as duplicate date entries and mix the types of the values in an array of Timestamps.

Timestamps are an immutable type, simply

A conversion method is provided to make a TimeArray into an Array{Timestamp}.

julia> using MarketData  # brings in TimeArray const objects

julia> ts = Array{Timestamp}(cl)

500-element Array{Timestamps.Timestamp{T},1}:

 2000-01-03 | 111.94

 2000-01-04 | 102.5  

 ⋮                  

 2001-12-28 | 22.43 

 2001-12-31 | 21.9  

With no checking of consistent value types, you might create an array with mixed types, and duplicate dates:

julia> nullable = Timestamp(today(), Nullable{Int}())

2015-03-11 | NA

julia> mystring = Timestamp(today(), "my string")

2015-03-11 | my string

julia> vcat(ts, nullable, mystring)

502-element Array{Timestamps.Timestamp{T},1}:

 2000-01-03 | 111.94   

 2000-01-04 | 102.5    

 ⋮                     

 2001-12-28 | 22.43    

 2001-12-31 | 21.9     

 2015-03-11 | NA       

 2015-03-11 | my string

Documentation is here: http://timestampsjl.readthedocs.org/en/latest/

Regards, 

Dan Wlasiuk