I'm working with the CausalImpact package for predicting a counterfactual time series. From prior analysis, I know that the true counterfactual series ought to show a downward trend, but the package seems to predict horizontal trends.

In other words, if I were to draw a line through the average value of the predicted counterfactual, it'll be a horizontal line. However, I have the prior knowledge it actually should be a line with negative slope. The supplied covariates actually show a slight downward trend as well, but apparently it's not strong enough for the package to pick up on.

Question: Is there a way to incorporate prior knowledge about trend in the CausalImpact package? I'm vaguely aware that bsts allows for a custom model but I don't know if it's doable.


closed as off-topic by jbowman, Jake Westfall, Michael Chernick, Peter Flom Feb 1 '18 at 13:18

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "This question appears to be off-topic because EITHER it is not about statistics, machine learning, data analysis, data mining, or data visualization, OR it focuses on programming, debugging, or performing routine operations within a statistical computing platform. If the latter, you could try the support links we maintain." – jbowman, Jake Westfall, Michael Chernick, Peter Flom
If this question can be reworded to fit the rules in the help center, please edit the question.