Contrib, and user supplied connection matrices.

[russelljjarvis]

I decided to restart the conversation here, since I realized I was no longer discussing an issue or a code break.

I made a PR, that shows how using the latest code changes, I can get pretty close to using plugin style connectivity matrices.
https://github.com/SpikingNetwork/TrainSpikingNet.jl/pull/4/files
StaticMatrices.jl almost work, PlasticWeightMatrices fail.

I wonder if it is possible to opt out of plastic simulations, by providing an empty Plastic weight Matrix?

The entry point to the code I provided is this:

TrainSpikingNet.jl/src/contrib$ julia configAndRunPotjans.jl 

This could also be regarded as a draft PR too. I might update it later.

Inside my code where I build the connection matrix, I save the matrix but I need to probably save it in the data-dir that the program user specifies, instead I save it in the pwd which might be src/contrib.

I should update the code so any matrices are stored in the user specified data_dir.

These are some files I changed:
https://github.com/russelljjarvis/TrainSpikingNet.jl/blob/master/src/contrib/Potjans_data/param.jl#L37-L55

I probably should convert the sparse matrices to dense before returning them.
https://github.com/russelljjarvis/TrainSpikingNet.jl/blob/connectomes/src/contrib/genPotjansConnectivity.jl#L244

julia> include("ConfigAndRunPotjans.jl")
LBTConfig([ILP64] libopenblas64_.so)
┌ Warning: running single threaded
└ @ TrainSpikingNet ~/build_testing/depth1/TrainSpikingNet.jl/src/paramfn.jl:57
p.genStaticWeights_file = "/home/rjjarvis/build_testing/depth1/TrainSpikingNet.jl/src/contrib/genPotjansConnectivity.jl"
Progress: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| Time: 0:00:11
┌──────────────────────────────┐
1 │⣿⣿⣿⣿⢏⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│ > 0
│⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│ < 0
│⣀⣀⣸⣿⣿⣿⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⢀⣀⠀⠀⠀⣀│
│⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⠉⠉⢹⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣋⡉⠉⠉⠉⠉⠉│
│⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⠰⡅⠀⠀⠀⠀⠀│
│⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡋⡅⠀⠀⠀⠀⠀│
│⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡔⠂⠀⠀⠀⠀⠀│
│⣶⣶⣾⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣶⣷⣶⣶⣶⣶⣶│
│⣿⣿⣿⣿⠟⡻⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⣿⣿⣿⣿⢽⣻⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
2 577 │⣿⣿⣿⣿⡷⢓⣿⣿⣿⣿⣿⣿⣿⣿⣿⣽⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
└──────────────────────────────┘
⠀1⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀2 577⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀557 316 ≠ 0⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
┌──────────────────────────────┐
1 │⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│ > 0
│⠀⠀⠀⠀⠀⠈⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│ < 0
│⠀⠀⠀⠀⠀⠀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⢀⣀⠀⠀⠀⣀│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣋⡉⠉⠉⠉⠉⠉│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⠰⡅⠀⠀⠀⠀⠀│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡋⡅⠀⠀⠀⠀⠀│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡔⠂⠀⠀⠀⠀⠀│
│⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣶⣷⣶⣶⣶⣶⣶│
│⠀⠀⠀⠀⠀⠈⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
│⠀⠀⠀⠀⠀⠈⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
2 577 │⠀⠀⠀⠀⠀⠀⣿⣿⣿⣿⣿⣿⣿⣿⣿⣽⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿│
└──────────────────────────────┘
⠀1⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀2 577⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀508 004 ≠ 0⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
┌──────────────────────────────┐
1 │⣿⣿⣿⣿⢏⣏⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│ > 0
│⣿⣿⣿⣿⣿⣷⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│ < 0
│⣀⣀⣸⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⣿⣿⣿⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⣿⣿⣿⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⣿⣿⣿⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⣿⣿⣿⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⠉⠉⢹⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⠀⠀⢸⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⠀⠀⢸⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⠀⠀⢸⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⣶⣶⣾⣿⣿⣿⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⣿⣿⣿⣿⠟⡻⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
│⣿⣿⣿⣿⢽⣳⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
2 577 │⣿⣿⣿⣿⡷⢓⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀│
└──────────────────────────────┘
⠀1⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀2 577⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀49 315 ≠ 0⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
ERROR: LoadError: BoundsError: attempt to access 2577-element Vector{Float32} at index [0]
Stacktrace:
[1] getindex
@ ./array.jl:924 [inlined]
[2] loop_init(itask::Int64, learn_every::Nothing, stim_on::Nothing, stim_off::Float64, train_time::Float64, dt::Float64, Nsteps::Int64, Ncells::Int64, Ne::Int64, Lei::Nothing, LX::Int64, refrac::Float64, learn_step::Int64, invtau_bale::Float32, invtau_bali::Float32, invtau_plas::Nothing, X_bal::Vector{Float32}, maxTimes::Nothing, times::Nothing, ns::Vector{UInt32}, timesX::Nothing, nsX::Vector{UInt32}, inputsE::Vector{Float32}, inputsI::Vector{Float32}, inputsP::Nothing, inputsEPrev::Vector{Float32}, inputsIPrev::Vector{Float32}, inputsPPrev::Nothing, spikes::Nothing, spikesPrev::Nothing, spikesX::Nothing, spikesXPrev::Nothing, u_bale::Vector{Float32}, u_bali::Vector{Float32}, uX_plas::Nothing, u_bal::Vector{Float32}, u::Vector{Float32}, r::Nothing, rX::Nothing, X::Vector{Float32}, wid::Nothing, example_neurons::Nothing, lastSpike::Vector{Float64}, plusone::Nothing, exactlyzero::Nothing, PScale::Nothing, raug::Nothing, k::Nothing, v::Vector{Float32}, rng::Random.TaskLocalRNG, noise::Vector{Float32}, rndX::Vector{Float32}, sig::Vector{Float32}, P::Nothing, w0Index::SparseArrays.SparseMatrixCSC{Int64, Int64}, w0Weights::SparseArrays.SparseMatrixCSC{Float32, Int64}, nc0::Vector{Int64}, X_stim::Nothing, utarg::Nothing, wpIndexIn::Nothing, wpIndexOut::Nothing, wpIndexConvert::Nothing, wpWeightX::Nothing, wpWeightIn::Nothing, wpWeightOut::Nothing, ncpIn::Nothing, ncpOut::Nothing, uavg::Vector{Float64}, ustd::Matrix{Float64}, rateX::Matrix{Float64}, cellModel_args::NamedTuple{(:thresh, :invtau_mem, :vre, :dt), Tuple{Vector{Float32}, Vector{Float32}, Float64, Float64}})
@ TrainSpikingNet ~/build_testing/depth1/TrainSpikingNet.jl/src/cpu/loop.jl:248
[3] init(; itasks::Vector{Int64}, utarg_file::Nothing, spikerate_file::Nothing)
@ TrainSpikingNet ~/build_testing/depth1/TrainSpikingNet.jl/src/initfn.jl:9
[4] init()
@ TrainSpikingNet ~/build_testing/depth1/TrainSpikingNet.jl/src/initfn.jl:1
[5] top-level scope
@ ~/build_testing/depth1/TrainSpikingNet.jl/src/contrib/ConfigAndRunPotjans.jl:40
[6] include(fname::String)
@ Base.MainInclude ./client.jl:476
[7] top-level scope
@ REPL[1]:1
[8] top-level scope
@ ~/.julia/packages/CUDA/BbliS/src/initialization.jl:52
in expression starting at /home/rjjarvis/build_testing/depth1/TrainSpikingNet.jl/src/contrib/ConfigAndRunPotjans.jl:40

[bjarthur]

I wonder if it is possible to opt out of plastic simulations, by providing an empty Plastic weight Matrix?

yes it is. or just use the default genPlasticWeights as you mentioned here: #1 (comment)

[bjarthur]

note that the adjacency matrices are now vectors of vectors instead of ragged dense matrices. see 2282152

[russelljjarvis]

Since making this PR I refactored the same sort of code in other places to make it faster and more memory efficient, using staticArrays.

https://github.com/russelljjarvis/SpikingNeuralNetworks.jl/blob/master/examples/genPotjansWiring.jl

I can wrangle the output static arrays into Vec{Vec} like in ()[https://github.com/SpikingNetwork/TrainSpikingNet.jl/commit/22821525fc62458f21f0936c36ac510b90ab5fb7], I will just have to wait for the right time.

[russelljjarvis]

I am also ultimately converting these matrices to {Vector{Vector{UInt32}}} so re integrating the divergent codes should be easy.