Description
The model coefficients could not be estimated due to lack of convergence, meaning that the estimates were not stable. This can have many causes, including extreme outliers, large data values, or high correlation between the explanatory variables. Other warning messages may indicate the source of this instability.
Solution
Review any additional warning messages and follow any recommendations, such as removing some explanatory variables or combining some categories.
Additionally, any of the following may resolve the error:
- Use Data Engineering to identify and remove any outliers in the data.
- Rescale any fields containing very large values, such as dividing all values by 1000.