NicoMosty / CellAggregate.jl

A julia implementation of some of the parallel agent based using CUDA for paralleling.

CellAggregate.jl

A julia implementation of some of the parallel agent based using CUDA for paralleling for Centre Model for Fusing Cell Aggregates.

Prerequisites

Install the next packages in julia

julia $ ]
    add CUDA
    add Adapt
    add ProgressMeter
    add DelimitedFiles
    add Images

Visualization

For a better visualization, its recommended the use of the ParaView or VMD program of the images in the data folder or test folder. The files are in the .xyz format.

Example

Model

julia $ @time model = ModelSet(
    TimeModel(
        tₛᵢₘ  = 150000.0,
        dt    = 0.5,
        nₖₙₙ  = 100,
        nₛₐᵥₑ = 50
    ),
    InputModel(
        outer_ratio = 0.8,
        path_input  = "../../data/init/Sphere"
    ),
    OutputModel(
        name_output = "Test_1",
        path_output = ""
    ) 
)

Aggregate

Struct

julia $ agg = Aggregate(
    [AggType(
        "Example", 
        InteractionPar(Par1, Par2),
        Float32.(readdlm("\$PATH")[3:end,2:end]) |> cu
    )], 
    [AggLocation("Example",[0 0 0]),],
    model
)

Fusion of Aggregates

Struct

julia $ fusion_agg = Aggregate(
    init_set,
    [
        AggLocation(init_set[1].Name,[-radius_loc[1] 0  0]),
        AggLocation(init_set[1].Name,[ radius_loc[1] 0  0])
    ],
    model
)

Complex Aggregates

Struct

julia $ pos = vcat([[2r*cos(i) 2r*sin(i)] for i=collect(0:60:360)*pi/180]...)
julia $ agg = Aggregate(
    [AggType(
        "HEK_1", 
        InteractionPar(Par1, Par2),
        init |> cu
    )], 
    [AggLocation("HEK_1",vcat(pos[i,:],0)') for i=1:size(pos,1)],
    model
)

Citation

Pending

Licence

Pending