Gnimuc

type Chain
  value :: Int
  right :: Chain
  left :: Chain
  #Make the last link in the chain point to itself
  #so as to spare us from the julia workaround for nulls
  Chain(value::Int) = (chain = new(); chain.value = value; chain.right = chain; chain.left = chain; chain)
end

Gnimuc

# load dependency packages
using GLFW, ModernGL

# set up OpenGL context version
@osx_only const VERSION_MAJOR = 4    # it seems OSX will stuck on OpenGL 4.1.
@osx_only const VERSION_MINOR = 1

# initialize GLFW library, error check is already wrapped.
GLFW.Init()

Gnimuc

function foo(genconv, I, J)
    genconv[sub2ind(size(genconv),I,J)]
end
function bar(genconv, I, J)
    diag(sub(genconv, I, J))
end
function baz(genconv, I, J)
    [genconv[i,j] for (i,j) in zip(I, J)]
end

Gnimuc

using Dierckx
function foo()
  x = [0., 1., 2., 3., 4.]
  y = [-1., 0., 7., 26., 63.]
  spl = Spline1D(x, y)
  a = evaluate(spl, ones(10000))
  reshape(a,100,100)
end

function bar()

Gnimuc

function topology(s1::Tuple{Int64,Int64}, s2::Tuple{Int64,Int64}, s3::Tuple{Int64,Int64}, a::Tuple{Int64,Int64}, b::Tuple{Int64,Int64}, c::Tuple{Int64,Int64})
	ks1 = (s1[1]+a[1], s1[2]+a[2])
	ks2 = (s2[1]+b[1], s2[2]+b[2])
	ks3 = (s3[1]+c[1], s3[2]+c[2])
	dφ1 = ks2[1] - ks1[1]
	dr1 = s2[1] - s1[1]
	dφ2 = ks2[2] - ks3[2]
	dr2 = s2[2] - s3[2]
	dφ3 = ks2[2] - ks1[2]
	# dr3 = dr1

Gnimuc

Yum

shell> cat /etc/centos-release
CentOS Linux release 7.3.1611 (Core) 

julia> versioninfo()
Julia Version 0.5.0
Commit 3c9d753 (2016-09-19 18:14 UTC)
Platform Info:
  System: Linux (x86_64-pc-linux-gnu)

Gnimuc

_       _ _(_)_     |  A fresh approach to technical computing
  (_)     | (_) (_)    |  Documentation: http://docs.julialang.org
   _ _   _| |_  __ _   |  Type "?help" for help.
  | | | | | | |/ _` |  |
  | | |_| | | | (_| |  |  Version 0.6.0-dev.2842 (2017-02-17 17:00 UTC)
 _/ |\__'_|_|_|\__'_|  |  Commit ded2d87* (0 days old master)
|__/                   |  x86_64-w64-mingw32

julia> Pkg.build("Cxx")

Gnimuc

# this is actually cross-correlation
function toyconv(x::AbstractArray{T,3}, w::AbstractArray{T,4}, pad::Integer, _stride::Integer) where T<:Real
   xSpatialDims = size(x, 1, 2)
   kernelSpatialDims = size(w, 1, 2)
   channelNum, kernelNum = size(w, 3, 4)
   ySpatialDims = 1 .+ (xSpatialDims .+ 2pad .- kernelSpatialDims) .÷ _stride
   δ = kernelSpatialDims[1] ÷ 2
   padSpatialDims = xSpatialDims .+ 2pad
   xOffset = OffsetArray(x, pad, pad, 0)
   xPadded = PaddedView(zero(T), xOffset, (padSpatialDims..., channelNum))

Gnimuc

        - module Hungarian
        - 
        - 
        - # Zero markers used in hungarian algorithm
        - # 0 => NON   => Non-zero
        - # 1 => Z     => ordinary zero
        - # 2 => STAR  => starred zero
        - # 3 => PRIME => primed zero
        - const NON = Int8(0)
        - const Z = Int8(1)

Gnimuc

        - """
        -     munkres(costMat) -> Zs
        - 
        - Find an optimal solution of the assignment problem represented by the square
        - matrix `costMat`. Return an sparse matrix illustrating the optimal matching.
        - 
        - # Examples
        - 
        - ```julia
        - julia> costMat = ones(3, 3) - eye(3, 3)