Calculates coefficients of predictor variables using real customer data (a point customer layer with an assigned store ID).
To calibrate the model's parameters, follow these steps:
- Obtain a map of the general area to be analyzed.
- Define the study area so it includes the trade areas of the stores being analyzed and the arterial networks of all competing stores.
- Divide the study area into subareas from which consumer patronage patterns can be observed. These subareas or origins should be small enough that each is relatively homogeneous in terms of the socioeconomic characteristics of a sufficient number to ensure shopping diversity.
You must have a point customers layer with corresponding shopping centers assigned. With this data, you can calculate the probability of consumers patronizing the competing stores and consequently resolve the equation.
You must split your study area into subareas to estimate the probability of customers from those areas patronizing a specific shopping center.
With a point layer containing customers with assigned corresponding store ID, you can complete the required estimation. You must divide your study area in such a way that at least several subareas contain customers patronizing different stores.
For example, if you have block groups dividing your study area so that all customers from each subarea are assigned to one shopping center, what does it mean? It means that the probability of customers from a specific subarea patronizing shopping centers will always equal either one or zero. To resolve the equation, the division into subareas must be performed in such a way that the estimated probability for at least several subareas is greater than zero and less than one.
Feedback on this topic?