Kernel CI Test for Relational Data (Lee and Honavar, UAI 2017), which accounts relational structure induced heterogeneity
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
uai2017experiments
.gitignore
README.md
requirements.txt
setup.py
version.py

README.md

A Kernel Relational Conditional Independence Test

Installation

If you are using macOS, wget is required, which can be installed using homebrew.

/usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
brew install wget

Other than standard python packages (e.g., numpy, scipy, matplotlib, etc), there are dependent repositories to be installed -- pyGK for graph kernels, pyRCDs for the implementation of relational causal model (RCM), and SDCIT for a suite of kernel-based (un)conditional independence tests.

Furthermore, tensorflow and GPflow are required. An easy way to install them is using anaconda and a separate environment.

cd ~/anaconda/bin
./conda create --name your_env_name python=3.6 --yes
source activate your_env_name
conda install six numpy wheel scipy matplotlib pandas  --yes 
conda install -c conda-forge tensorflow --yes
cd ~/Downloads
git clone https://github.com/GPflow/GPflow
git clone https://github.com/sanghack81/pyGK
git clone https://github.com/sanghack81/SDCIT
git clone https://github.com/sanghack81/pyRCDs
git clone https://github.com/sanghack81/KRCIT
cd GPflow
python3 setup.py install
cd ../SDCIT
conda install --yes --file requirements.txt
./setup.sh
python3 setup.py install
cd ../pyGK
conda install --yes --file requirements.txt
python3 setup.py install
cd ../pyRCDs
conda install --yes --file requirements.txt
python3 setup.py install
cd ..
rm -rf SDCIT pyGK pyRCDs GPflow KRCIT

Test whether all required packages are installed correctly.

import tensorflow
import GPflow
import pygk
from sdcit.sdcit import SDCIT
import pyrcds