SpecIV/SpecPCL

Spectral Representation for Causal Estimation with Hidden Confounders

Requirements

Please see the requirements.txt file for the required packages.

datasets/ contains the datasets we will use in the experiments.
- demand_design.py generates the synthetic demand design dataset.
- dsprite.py generates the synthetic dSprites dataset.
algos/ contains the implementation of the algorithms.
- embedding_sgd.py implements the dual embedding SGD algorithm, i.e., the vanilla version of our method.
- spectral_repr.py implements the Spectral Representation for Causal Inference algorithm.
configs contains the configuration files for the experiments. In each yaml file:
networks/ contains the implementation of the neural network parametrizations used in the experiments.
- image_models.py implements the CNN for feature extraction.
- contrastive_models.py implements the MLP for the contrastive loss-based model (CTRL).
utils/ contains the implementation of the utility functions.
- dist.py implements the distance functions.
- lin_alg.py implements the linear algebra functions.
- kernel.py implements the kernel constructing functions.
- sample_generator.py implements the sample generating functions used in the embedding SGD experiments.
logs/ contains the logs of the experiments.
main.py is the entrance of the algorithm.

To run the code, you can use the following command:

python main.py --config <path-to-configs>