DEEP_causal
Project description
DEEP_causal
DEEP_causal is a Python library for discovering causal heterogeneity. The mechanism can be found in the publication "Causal heterogeneity discovery by bottom-up pattern search".
Installation
Use the package manager pip to install DEEP_causal. Install_requires=["gsq==0.1.6", "pandas==1.3.4", "numpy==1.21.4"]
pip install gsq==0.1.6
pip install pandas==1.3.4
pip install numpy==1.21.4
pip install DEEP_causal
Usage
from DEEP_causal import compute_DEEP_causal
import pandas as pd
# read example training data and testing data
# example data is available at https://drive.google.com/drive/folders/1UW_bcZ7d_ygTXRjJlfR4JrvFKJwWq30D
train_data = pd.read_csv("example_train.csv")
test_data = pd.read_csv("example_test.csv")
# parents, treatment, outcome
parents_trtout = ["x1", "x2", "x3", "x4", "x5", "x6", "x7", "x8", "w", "y"] # from MMPC
# returns patterns and causal effect
train_output, test_output = compute_DEEP_causal.compute_dc(train_data, parents_trtout, test_data)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
DEEP_causal-0.0.22.tar.gz
(5.4 kB
view hashes)