I\'ve built a linear regression model predicting recidivism among convicts based on the COMPAS dataset. I\'ve some issues regarding the categorical variables, specifically the g