Dynamical Wasserstein Barycenters for Time Series Modeling
This is the code related for the Dynamical Wasserstein Barycenter model published in Neurips 2021.
To run the code and replicate the results reported in our paper,
# usage: DynamicalWassersteinBarycenters.py dataSet dataFile debugFolder interpModel [--ParamTest PARAMTEST] [--lambda LAM] [--s S]
# Sample run on MSR data
>> python DynamicalWassersteinBarycenters.py MSR_Batch ../Data/MSR_Data/subj090_1.mat ../debug/MSR/subj001_1.mat Wass
# Sample run for parameter test
>> python DynamicalWassersteinBarycenters.py MSR_Batch ../Data/MSR_Data/subj090_1.mat ../debug/ParamTest/subj001_1.mat Wass --ParamTest 1 --lambda 100 --s 1.0
The interpMethod is either Wass` for the Wasserstein barycentric model or GMM`` for the linear interpolation model.
Simulated Data
The simulated data and experiment included in this supplement can be replicated using using the following commands.
# Generate 2 and 3 state simulated data
>> python GenerateOptimizationExperimentData.py
>> python GenerateOptimizationExperimentData_3K.py
# usage: OptimizationExperiment.py FileIn Mode File
# Sample run for optimization experiment
>> python OptimizationExperiment.py ../data/SimulatedOptimizationData_2K/dim_5_5.mat/ WB ../debug/SimulatedData/dim_5_5_out.mat
The Mode is either WB for Wasserstein-Bures geometry and Euc for Euclidean geometry using Cholesky decomposition parameterization.
Requirements
_libgcc_mutex=0.1=conda_forge
_openmp_mutex=4.5=1_llvm
_pytorch_select=0.2=gpu_0
blas=2.17=openblas
ca-certificates=2020.12.5=ha878542_0
certifi=2020.12.5=py38h578d9bd_1
cffi=1.14.4=py38h261ae71_0
cudatoolkit=8.0=3
cudnn=7.1.3=cuda8.0_0
cycler=0.10.0=py_2
freetype=2.10.4=h7ca028e_0
future=0.18.2=py38h578d9bd_3
immutables=0.15=py38h497a2fe_0
intel-openmp=2020.2=254
joblib=1.0.0=pyhd8ed1ab_0
jpeg=9d=h36c2ea0_0
kiwisolver=1.3.1=py38h82cb98a_0
lcms2=2.11=hcbb858e_1
ld_impl_linux-64=2.33.1=h53a641e_7
libblas=3.8.0=17_openblas
libcblas=3.8.0=17_openblas
libedit=3.1.20191231=h14c3975_1
libffi=3.3=he6710b0_2
libgcc-ng=9.3.0=h5dbcf3e_17
libgfortran-ng=7.3.0=hdf63c60_0
libgomp=9.3.0=h5dbcf3e_17
liblapack=3.8.0=17_openblas
liblapacke=3.8.0=17_openblas
libopenblas=0.3.10=pthreads_hb3c22a3_4
libpng=1.6.37=h21135ba_2
libstdcxx-ng=9.3.0=h6de172a_18
libtiff=4.1.0=h4f3a223_6
libwebp-base=1.1.0=h36c2ea0_3
llvm-openmp=11.0.0=hfc4b9b4_1
lz4-c=1.9.2=he1b5a44_3
matplotlib-base=3.3.3=py38h5c7f4ab_0
mkl=2020.4=h726a3e6_304
mkl-service=2.3.0=py38he904b0f_0
mkl_fft=1.3.0=py38h5c078b8_1
mkl_random=1.2.0=py38hc5bc63f_1
ncurses=6.2=he6710b0_1
ninja=1.10.2=py38hff7bd54_0
numpy=1.19.5=py38h18fd61f_1
numpy-base=1.18.5=py38h2f8d375_0
olefile=0.46=pyh9f0ad1d_1
openssl=1.1.1k=h7f98852_0
pillow=8.1.0=py38h357d4e7_1
pip=20.3.3=py38h06a4308_0
pot=0.7.0=py38h950e882_0
pycparser=2.20=py_2
pyparsing=2.4.7=pyh9f0ad1d_0
python=3.8.5=h7579374_1
python-dateutil=2.8.1=py_0
python_abi=3.8=1_cp38
pytorch=1.7.1=cpu_py38h36eccb8_1
readline=8.0=h7b6447c_0
scikit-learn=0.24.1=py38h658cfdd_0
scipy=1.5.2=py38h8c5af15_0
setuptools=51.1.2=py38h06a4308_4
six=1.15.0=py38h06a4308_0
sqlite=3.33.0=h62c20be_0
threadpoolctl=2.1.0=pyh5ca1d4c_0
tk=8.6.10=hbc83047_0
tornado=6.1=py38h497a2fe_1
wheel=0.36.2=pyhd3eb1b0_0
xz=5.2.5=h7b6447c_0
zlib=1.2.11=h7b6447c_3
zstd=1.4.5=h6597ccf_2
4.5k Jan 07, 2023
2k Jan 05, 2023
158 Dec 25, 2022
16 Dec 13, 2022
5 Jun 08, 2022
213 Nov 12, 2022
13 Nov 24, 2022
89 Jan 05, 2023
97 Dec 21, 2022
87 Nov 29, 2022
2 Jul 01, 2022
21 Dec 19, 2022
261 Nov 12, 2022
7 Oct 31, 2021
33 Nov 30, 2022
21 Nov 26, 2022
1.2k Dec 25, 2022
134 Jan 03, 2023
305 Jan 03, 2023
17 Nov 22, 2022