Giter Club home page Giter Club logo

ucate's Introduction

Introduction

This repository contains the code to replicate the results reported in Identifying Causal Effect Inference Failure with Uncertainty-Aware Models.

Intuitive Uncertainty for CATE

Installation

$ git clone [email protected]:OATML/ucate.git
$ cd ucate
$ conda env create -f environment.yml
$ conda activate ucate

Download data

$ mkdir data

To run the following experiments, the IHDP train, test, and ACIC2016 datasets must be downloaded into the folder created above. The datasets will need to be uncompressed.

BTLearner Experiment

IHDP

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    tlearner
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tlearner/bf-200_dp-5_dr-0.5_bs-100_lr-0.001_ep-False

IHDP Covariate Shift

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    --exclude-population True
    tlearner
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tlearner/bf-200_dp-5_dr-0.5_bs-100_lr-0.001_ep-False

CEMNIST

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name cemnist \
    --data-dir data/ \
    --num-trials 20
    tlearner
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/cemnist/tlearner/bf-200_dp-5_dr-0.5_bs-100_lr-0.001_ep-False

ACIC

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name acic \
    --data-dir data/data_cf_all/ \
    --num-trials 77
    tlearner
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/acic/tlearner/bf-200_dp-5_dr-0.5_bs-100_lr-0.001_ep-False

BTARNet Experiment

IHDP

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    tarnet
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tarnet/md-tarnet_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

IHDP Covariate Shift

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    --exclude-population True
    tarnet
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tarnet/md-tarnet_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-True

CEMNIST

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name cemnist \
    --data-dir data/ \
    --num-trials 20
    tarnet
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/cemnist/tarnet/md-tarnet_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

ACIC

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name acic \
    --data-dir data/data_cf_all/ \
    --num-trials 77
    tarnet
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/acic/tarnet/md-tarnet_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

BCFR-MMD Experiment

IHDP

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    tarnet \
    --mode mmd
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tarnet/md-mmd_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

IHDP Covariate Shift

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    --exclude-population True
    tarnet \
    --mode mmd
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tarnet/md-mmd_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-True

CEMNIST

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name cemnist \
    --data-dir data/ \
    --num-trials 20
    tarnet \
    --mode mmd
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/cemnist/tarnet/md-mmd_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

ACIC

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name acic \
    --data-dir data/data_cf_all/ \
    --num-trials 77
    tarnet \
    --mode mmd
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/acic/tarnet/md-mmd_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

BDragonnet Experiment

IHDP

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    tarnet \
    --mode dragon
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tarnet/md-dragon_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

IHDP Covariate Shift

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    --exclude-population True
    tarnet \
    --mode dragon
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/tarnet/md-dragon_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-True

CEMNIST

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name cemnist \
    --data-dir data/ \
    --num-trials 20
    tarnet \
    --mode dragon
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/cemnist/tarnet/md-dragon_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

ACIC

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name acic \
    --data-dir data/data_cf_all/ \
    --num-trials 77
    tarnet \
    --mode dragon
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/acic/tarnet/md-dragon_bf-200_dr-0.5_beta-1.0_bs-100_lr-0.001_ep-False

BCEVAE Experiment

IHDP

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    cevae
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/cevae/dl-32_bf-200_dr-0.1_beta-0.1_ns-True_bs-100_lr-0.001_ep-False

IHDP Covariate Shift

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    --exclude-population True
    cevae
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/cevae/dl-32_bf-200_dr-0.1_beta-0.1_ns-True_bs-100_lr-0.001_ep-True

CEMNIST

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name cemnist \
    --data-dir data/ \
    --num-trials 20
    cevae
    --learning-rate 2e-4
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/cemnist/cevae/dl-32_bf-200_dr-0.1_beta-0.1_ns-True_bs-100_lr-0.0002_ep-False

ACIC

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name acic \
    --data-dir data/data_cf_all/ \
    --num-trials 77
    cevae
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/acic/cevae/dl-32_bf-200_dr-0.1_beta-0.1_ns-True_bs-100_lr-0.001_ep-False

BART Experiment

IHDP

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    bart
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/bart/ep-False

IHDP Covariate Shift

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name ihdp \
    --data-dir data/ \
    --num-trials 1000 \
    --exclude-population True
    bart
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/ihdp/bart/ep-True

CEMNIST

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name cemnist \
    --data-dir data/ \
    --num-trials 20
    bart
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/cemnist/bart/ep-False

ACIC

$ ucate train \
    --job-dir ~/experiments/ucate \
    --dataset-name acic \
    --data-dir data/data_cf_all/ \
    --num-trials 77
    bart
$ ucate evaluate \
    --experiment-dir ~/experiments/ucate/acic/bart/ep-False

ucate's People

Contributors

anndvision avatar hadarrosenwald avatar

Watchers

James Cloos avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.