LisaMaag/QAS: Quasi-analytical solution for logit

The package provides a quasi-analytical solution for logit models. Numerical optimization for logit models is often time consuming, at least for big data. This package avoids this problem with a quasi-analytical approach. The resulting coefficients are especially suitable for imputation models, but they differ from GLM results. Therefore the coefficients are less suitable for statistical modelling and interpretation. The predictQAS-Function can be used after the main QAS.func-Function to categorize numeric variables in the same way as the QAS.func-Function for calculating the coefficients. The function can also predict values for the dependent variable based on the coefficients of QAS.func and the dataset used for categorization.

install.packages("remotes")
remotes::install_github("LisaMaag/QAS")