Jamal Senouci

breaking out the wider confidence intervals from the rest of the issues raised in #7. This is potentially due to the R version placing an [upper limit](https://github.com/google/CausalImpact/blob/master/R/impact_model.R) on the standard...