Search code examples
pythonpandaspredictpredictive

Interpreting the probability of still being alive matrix plot BG/NBD model


I am implementing a BG/NBD model and have plotted the probability of being alive matrix, which is the following:

fig = plt.figure(figsize=(15,10))
plot_probability_alive_matrix(bgf)
plt.show()

enter image description here

Can someone help me interpret what the vertical yellow line at the left of the plot is? I know that bottom right means a customer is most likely "alive", and top right is where customers who have most likely dropped out are. But what is the meaning of the yellow bar at the left?

Thanks!


Solution

  • This is normal. Costumer attrition can only occur after a purchase has been made. When none is done (frequency = 0), the costumer is "alive"and P(Alive) = 1.