Select simplest good model

Select simplest fit within one SE of best fit

select_cvlt(cvlt_fits)

Arguments

cvlt_fits	result of running fit_ldats_multiple

Value

dataframe with specifications for simplest model with loglikelihood within 1 SE of the loglikelihood of the best model

Details

First minimizes the number of changepoints, then the number of topics. Then selects the LDA seed with the best loglikelihood.