generate_unsaturated_data

pymc_marketing.customer_choice.synthetic_data.generate_unsaturated_data(total_sales_before, total_sales_after, total_sales_sigma, treatment_time, n_observations, market_shares_before, market_shares_after, market_share_labels, random_seed=None)

Generate synthetic data for the MVITS model.

Notably, we can define different total sales levels before and after the introduction of the new model.

This function generates synthetic data for the MVITS model, assuming that the market is unsaturated meaning that there are new sales to be made.

This makes the assumption that the total sales are normally distributed around some average level of sales, and that the market shares are constant over time.

Parameters:

total_sales_mu: int

The average level of sales in the market.

total_sales_sigma: float

The standard deviation of sales in the market.

treatment_time: int

The time at which the new model is introduced.

n_observations: int

The number of observations to generate.

market_shares_before: list[float]

The market shares before the introduction of the new model.

market_shares_after: list[float]

The market shares after the introduction of the new model.

market_share_labels: list[str]

The labels for the market shares.

random_seed: np.random.Generator | int, optional

The random number generator to use.

Returns:

data: pd.DataFrame

The synthetic data generated.