NonstationaryPhase

This is a Julia package which holds all the source code used in the paper "A generalized quadratic estimate for random field nonstationarity" by Ethan Anderes (University of California at Davis) and Joe Guinness (North Carolina State University).

All the code was run on Julia 0.4 (installation instructions can be found at julialang.org). The exact version information for the system which generated the figures are given as follows.

julia> Pkg.installed("PyPlot")
v"2.1.1"

julia> Pkg.installed("Distributions")
v"0.8.10"

julia> versioninfo()
Julia Version 0.4.4-pre+29
Commit d2a82a0 (2016-02-09 14:59 UTC)
Platform Info:
  System: Darwin (x86_64-apple-darwin15.3.0)
  CPU: Intel(R) Core(TM) i7-4850HQ CPU @ 2.30GHz
  WORD_SIZE: 64
  BLAS: libopenblas (USE64BITINT DYNAMIC_ARCH NO_AFFINITY Haswell)
  LAPACK: libopenblas64_
  LIBM: libopenlibm
  LLVM: libLLVM-3.3

This information is intended to enable anyone to recreate the exact system which generated the figures in the paper. That said, we expect the code to be forward compatible with stable releases.

Using Julia to automatically download the source code and the paper from Github

The following code uses the package manager in Julia to download and install the package NonstationaryPhase.

julia> Pkg.clone("https://github.com/EthanAnderes/NonstationaryPhase.jl.git")

Note: julia> above indicates that the code which follows should be executed in the Julia REPL (i.e. the Julia command line). Typically the package is downloaded to directory ~/.julia/v0.4/ or something similar. You can also use Julia to tell you where the package is with the command Pkg.dir("NonstationaryPhase") in the Julia REPL.

Now use the following code to generate Figure 1 (for example) from the paper.

julia> using NonstationaryPhase 
julia> figure1_path = joinpath(Pkg.dir("NonstationaryPhase"),"figures/make_figure_1.jl")
julia> include(figure1_path) 

This will save figure1.pdf to the directory that the Julia REPL was launched.

Removing the package

You can either directly delete the package directory (e.g. use something like rm -fr ~/.julia/v0.4/NonstationaryPhase in the terminal). Or you can use the Julia package manager to remove it for you with the command Pkg.rm("NonstationaryPhase").

Clone the package directly from Github.

If you want to avoid using the Julia package manager you can use the following code to download the repo manually from the terminal.

$ git clone https://github.com/EthanAnderes/NonstationaryPhase.jl.git

You need to make sure Julia knows where to find NonstationaryPhase.jl.

julia> push!(LOAD_PATH, "<path to the directory containing NonstationaryPhase.jl>")

Now you should be able to run the following code to generate Figure 1 (for example).

julia> using NonstationaryPhase
julia> include("<path to NonstationaryPhase.jl>/figures/make_figure_1.jl")

This will save figure1.pdf to the directory that the Julia REPL was launched.

Runtime

When running the code be sure to keep in mind that figures 2, 3 and 4 are simulation heavy. Therefore, expect them to take a few hours to run (but no longer than one day on a standard laptop).

Bug fixes, revisions and pull requests

We plan to experiment with keeping the paper "alive" by continually adding bug fixes and revisions, even after possible publication. We will also consider accepting pull requests for bug-fixes and revisions but only if/when publication occurs to avoid co-author conflicts. Note that pull requests can all be done within the Julia package manager (instructions can be found here).