You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I tried to implement the code with the dataset you provided: dt_simulated_weekly and dt_prophet_holidays, with same hyperparameter settings and without calibration settings. As for the model, I selected one from Pareto-optimals, 5_277_3.
However, I noticed that in the allocation result, for Facebook_S, the mean response shows 0% despite the date range. You can see that in the first table, the date range is about 2 years, and in the second the date range is 4 weeks.
This is expected. The selected model has predicted facebook_S has no effect / 0 coefficient (see grey table 1st row 3rd column), thus it'll always get 0 allocation. pick another model that has no 0 coef for any channels, you'll see "normal result".
Hi Robyn team, thanks for this amazing tool.
I tried to implement the code with the dataset you provided: dt_simulated_weekly and dt_prophet_holidays, with same hyperparameter settings and without calibration settings. As for the model, I selected one from Pareto-optimals, 5_277_3.
However, I noticed that in the allocation result, for Facebook_S, the mean response shows 0% despite the date range. You can see that in the first table, the date range is about 2 years, and in the second the date range is 4 weeks.
Here is the JSON file for Robyn_Model-5_277_3:
RobynModel-5_277_3.json
I'd appreciate it if there's any response on this issue.
ps.
Robyn version: 3.11.1
R version: 4.3.1
The text was updated successfully, but these errors were encountered: