Computing shares after a change in exogenous variable about pyblp HOT 5 CLOSED

from pyblp.

(Added in 51433dc)

from pyblp.

I think you've almost got it. You should definitely be using the Simulation class when doing counterfactuals that aren't explicitly supported by ProblemResults.

To keep prices the same, pass the prices you want to the prices argument of Simulation.replace_endogenous and set iteration=Iteration('return'). This just uses the iteration "routine" that simply returns the the starting values, which are prices. You can verify that prices haven't changed by looking at the product_data attribute of the returned simulation results.

from pyblp.

Jeff, can you add this to the documentation somewhere?

To keep prices the same, pass the prices you want to the prices argument of Simulation.replace_endogenous and set iteration=Iteration('return'). This just uses the iteration "routine" that simply returns the the starting values, which are prices. You can verify that prices haven't changed by looking at the product_data attribute of the returned simulation results.

from pyblp.

Yeah for sure, good point.

from pyblp.