有效地使用具有多个模型的predict.glm

I am weighing the efficacy of using one monolithic model, versus splitting out in to two different models (a split model) on about 100,000 rows of data. To do so, I am getting results from my split model like so:I am weighing the efficacy of using one monolit