fcdimitr / sgtsnepi.jl

Julia bindings to SG-t-SNE-Π

We provide a Julia interface, i.e., a wrapper to SG-t-SNE-Π, which is a high-performance software for swift embedding of a large, sparse graph into a d-dimensional space (d = 1,2,3) on a shared-memory computer.

Installation

To install SG-t-SNE-Π through Julia, issue

] add SGtSNEpi

⚠️ SGtSNEpi is currently not working on Windows and native M1 Macs: Either use WSL2 on Windows or use the package via Rosetta 2 on the M-processor Macs (that means by using the x86 and not the arm64 julia).

See the full documentation for more details.

Citation

If you use this software, please cite the following paper:

@inproceedings{pitsianis2019sgtsnepi,
    author = {Pitsianis, Nikos and Iliopoulos, Alexandros-Stavros and Floros, Dimitris and Sun, Xiaobai},
    doi = {10.1109/HPEC.2019.8916505},
    booktitle = {IEEE High Performance Extreme Computing Conference},
    month = {11},
    title = {{Spaceland Embedding of Sparse Stochastic Graphs}},
    year = {2019}
}

sgtsnepi.jl's Issues

UndefVarError: libsgtsnepi not defined

Hi there,

I was following the Graph embedding guide in the documentation and there is an issue with SGtSNEpi.jl.

I ran this code after downloading the sample data in that the documentation references:

using DataDeps, MatrixMarket

L = vec( MatrixMarket.mmread( datadep"optdigits/optdigits_10NN/optdigits_10NN_label.mtx" ) );
L = Int.( L );

using SGtSNEpi, Random

# reproducible results
Random.seed!(0);
Y0 = 0.01 * randn( size(A,1), 2 );

Y = sgtsnepi(A; Y0 = Y0);

The program ran successfully up to line 12, where this error occurs:

ERROR: UndefVarError: libsgtsnepi not defined
Stacktrace:
 [1] _sgtsnepi_c(P::SparseArrays.SparseMatrixCSC{Float64, Int64}, d::Int64, max_iter::Int64, early_exag::Int64, λ::Int64; Y0::Matrix{Float64}, np::Int64, h::Float64, bb::Float64, eta::Float64, run_exact::Bool, fftw_single::Bool, alpha::Int64, profile::Bool, drop_leaf::Bool, list_grid_size::Vector{Int64}, par_scheme_grid_thres::Int64)
   @ SGtSNEpi C:\Users\d8amo\.julia\packages\SGtSNEpi\4wQHp\src\sgtsne.jl:250
 [2] sgtsnepi(A::SparseArrays.SparseMatrixCSC{Float64, Int64}; d::Int64, λ::Int64, max_iter::Int64, early_exag::Int64, Y0::Matrix{Float64}, profile::Bool, np::Int64, version::SGtSNEpi.SGTSNEPI_VERSION, h::Float64, u::Int64, k::Int64, eta::Float64, alpha::Int64, fftw_single::Bool, exact::Bool, drop_leaf::Bool, list_grid_size::Vector{Int64}, bound_box::Float64, par_scheme_grid_thres::Int64)
   @ SGtSNEpi .\array.jl:0
 [3] top-level scope
   @ c:\Users\path\to\my\file\SGtSNEpi_test.jl:12

Following the stack trace, the error occurs on line 250 of sgtsne.jl. Do you know how to fix this? Thanks.

FLANN dependancy issue

Hi,

Due to an issue on FLANN side (see wildart/FLANN.jl#25), this package do not work anymore :/

Would you have any idea for a work-around by any chance?

Negative weights on input graph

In case the input graph has negative weights, issue error.

Currently, the code silently fails, returning NaN values.

Example Code does not work on my machine

I have been exploring ways to display large but sparse networks and your algorithm came up. I have tried to use the sample code from the Graph plotting tutorial in Graphs.jl documentation (you can find that code here). This code does not appear to run, and the error that Julia is telling me is occuring is an undefined variable libsgtsnepi, which as far as I can tell is only used in the c call in sgtsne.jl.

TagBot trigger issue

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!