consumer_taste_grid#

pymc_marketing.customer_choice.taste_profiles.consumer_taste_grid(model)[source]#

Halton grid as a labelled DataFrame.

Parameters:
modelBayesianBLP

Constructed (need not be fitted) BayesianBLP.

Returns:
pd.DataFrame

Shape (n_mc_draws, n_random). Column names are the random-coefficient names, in the same order as model._random_coef_names. Each row is one consumer type; column d carries the standard-normal taste draw for that random-coefficient dimension.