Replies: 1 comment
-
|
Hello @njprice17 , Thank you for contacting us! You cannot use roi_m and beta_m priors for different channels within the same model instance as ModelSpec applies one prior type across all media channels. In Meridian, the PriorDistribution used depends on the media_prior_type. See here for more details. Therefore, we recommend proceeding with one type. Meridian ignores beta_m distribution if paid_media_prior_type is 'roi' or 'mroi', and uses the roi_m or mroi_m prior instead. Additionally, ROI priors are generally recommended for their ease of customization, direct alignment with business goals, and facilitation of experiment calibration. Coefficient priors should be avoided due to their uninformative nature and difficulty in customization, making them unsuitable for most applications, except when an extremely uninformative prior is specifically needed, which is not recommended. For guidance on selecting prior types, refer to our documentation on How to choose treatment prior types. Feel free to reach out for any further queries or feedback regarding Meridian. Google Meridian Support Team |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi, I'm currently trying to use more than one media type, lets say I have 8 channels and I want to use the roi_m prior for 1 and the rest as beta_m. I'm currently trying to do this but get an error saying the lists I'm passing must by 8 in length to match the number of media channels.
Could someone provide an example of how to do this?
Thanks!
Beta Was this translation helpful? Give feedback.
All reactions