Based on use by the SEMCOG team and conversation with @jpn-- , we'll make a few little improvements to the notebook examples: - add the number of iterations to the estimate call: m.estimate(method='SLSQP', options={'maxiter’:1000}) - speed up writing the Excel report: m.to_xlsx(“filename.xlsx", data_statistics=False) - set the av (availability) vector based on util size term > 0 - filter out choosers who don't make a valid observed choice - update the dependencies to use larch.util.activitysim instead of asim_larch.py for better packaging - add m.parameter_summary() to the notebook reporting so the user can see the t-stats - add text above the estimate call to explain when to use SLSQP versus BHHH
Activity
additional updates for estimation integration (#328)
Develop (#366)