tbeason / dailytreasuryyieldcurve.jl Goto Github PK

Glancing over the code again, I see some potential updates.

• No need to require DataFrames.jl when I could just use Tables.jl. A standard column table will probably suffice, and the docs can just recommend conversion to DataFrame if that is what is desired.

• Interpolation should be cleaned up.

• There are other data sources that might be better options given the annoying XML nature of the original source. https://www.financialresearch.gov/ is one alternative.

10-11-10 exists in the file yet has no rates given. Should just remove this date from the dataset altogether when constructing a rate interpolator.

The warning message that prints comes from EzXML. I've opened an issue there. JuliaIO/EzXML.jl#135

Things that could be done:

• date filters to download less (but easy to do after the fact already)
• add helper methods for building interpolators (linear/spline from Interpolations.jl?)
• actually add documentation