Skip to content

Collection of methods for instrumental variable settings with a compositional cause.

Notifications You must be signed in to change notification settings

bio-datascience/comp-iv

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 

Repository files navigation

A causal view on compositional data

This is the code for reconstructing the experiments in A Causal view on Compositional Data (E. Ailer, C.L. Müller and N. Kilbertus).

Setup

First, clone this repository.

git clone git@github.com:eailer/comp-iv.git

To run the code, please first create a new Python3 environment (Python version >= 3.7 should work). Then install the required packages into your newly created environment via

python -m pip install -r requirements.txt

Experiments

The experiments are explained step by step in individual jupyter notebooks.

cd notebooks

The experiments are separated in three different notebooks.

One Dimensional IV Estimation

The notebook 1_RealData_Diversityis centered around the motivational experiments. For input, it takes an instrumental variable setup with parameters Z, X, Y, whereas X is a diversity estimate.

Compositional IV Estimation

The other notebooks 2_MicrobiomeAnalysis_SimulatedData and 2_MicrobiomeAnalysis_RealData include the higher dimensional setup with compositional X. They include the exact parameter settings used in the paper. Moreover, they provide the possibility to exchange parameters or use own data to apply the proposed methods.

About

Collection of methods for instrumental variable settings with a compositional cause.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 99.4%
  • Python 0.6%