search

py-why / dowhy

DoWhy is a Python library for causal inference that supports explicit modeling and testing of causal assumptions. DoWhy is based on a unified language for causal inference, combining causal graphical models and potential outcomes frameworks.
https://www.pywhy.org/dowhy
MIT License
7.15k stars 935 forks source link

Consistently different values for estimated effect? #1238

Closed g-adje closed 2 months ago

g-adje commented 3 months ago

Hi - I am exploring a DAG that I have created from time series data by using causalnex's implementation of DYNOTEARS, taking into account about 10 features and their lag-1 equivalents.

A goal is to estimate the effect of changing a lag-1 feature on another lag-0 feature.

On data/DAG we have concocted, for which we know the ground truth, it all goes very well. However, when using identify_effect() and estimate_effect() on the data/DAG we actually care about, we are getting very different results each time we execute the function - not only does the mean value change considerably but so does the entire list of effect estimates.

I assume there is some theoretical point that I'm missing here that might prevent this from happening?

Very happy to provide more information if necessary. Thanks for any help you might be able to provide.

Version information: 0.9.1

github-actions[bot] commented 3 months ago

This issue is stale because it has been open for 14 days with no activity.

github-actions[bot] commented 2 months ago

This issue was closed because it has been inactive for 7 days since being marked as stale.