balance
balance copied to clipboard
facebookresearch
The balance python package offers a simple workflow and methods for dealing with biased data samples when looking to infer from them to some target population of interest.
Regarding balance/graviton (once we deal with the glmnet->sklearn transition). How about we just move to using max_de to be based on weight trimming only (instead of shrinkage in the LASSO...
Not sure yet if it's better to raise an error, or a warning (and impute it with 0s). But let's go with error. But at least it's worth giving a...
Using calude 3 opus In this rewritten version: The cvglmnet and related functions have been replaced with sklearn's LogisticRegression and GridSearchCV for model fitting and hyperparameter tuning. The choose_regularization function...
# Describe the bug The same code has no error when running method ='ipw', and method = 'cbps', but return below error when using raking. The below code return error...
This is violating user expectation. E.g.: ``` weights_untrimmed = sample.adjust( variables=weighting_variables, method="rake", weight_trimming_mean_ratio=0, transformations=auto_recodes, ) ``` By expectation violation I mean, 1. there isn’t anything in the docs that suggests...
Currently the usage of glmnet_python in ipw fuction is a blocker to: 1. Using ipw on Windows (due to installation of glment_pyton) (issue: https://github.com/facebookresearch/balance/issues/26) 2. Migrating the license from GPL2...
We suspect that when providing a feature that has mostly unique string values and some NaN values, then this might lead the ipw model to give just weights of 1....
Current output:  It could be better to make sure the lines of covars are next to each other. That the diagnostics of the weights include also the ESSP and...
After adjustment, the object doesn't show any information of interest about the adjustment used (e.g.: ipw, or Deff, etc.) Consider to update it a bit (at least for the statistics...
Formulaic is a high-performance implementation of Wilkinson formulas for Python. https://github.com/matthewwardrop/formulaic It's probably worth transitioning to it once it matures.
Metadata
Owner
Metadata
The balance python package offers a simple workflow and methods for dealing with biased data samples when looking to infer from them to some target population of interest.