Google Causal Impact Case Study in R

[@kofimawuli8940]

Great work on applying the causal impact tool to empirical data.

[@judygu9276]

Great work! Any reason to use adjusted price not returns ?

[@DataHeroes]

You can use any. Really depends what you are trying to predict

[@econrodrigoozon]

For the "fomat error" u can solve this fixing the pre.period by including one day after the treatment date.
Someting like this:
start

[@DataHeroes]

Thanks, Rodrigo, for thr suggestion

[@coloredmesurprised]

Fantastic work, really beneficial. One question: are there any other data package like series that can readily be used as a source of time series data? For example, import, export, GDP, etc.?

[@DataHeroes]

Unfortunately, none that I am familiar with

[@raiadscaggion9338]

Awesome! Keep up the good work, man! :)

[@DataHeroes]

Thanks!

[@knowledgeheightsacademy2466]

I keep getting 'data is not numeric,. What could be the problem?

[@knowledgeheightsacademy2466]

@@DataHeroes I added to my library "CausalImpact" and run the following;
> #Read the data
> B #Using a different model again
> start="2009-3-1"
> treatment="2014-12-1"
> end="2020-9-1"
> # define the pre and post periods
> pre.period=as.Date(c(start, treatment))
> post.period=as.Date(c(treatment, end))
> #Declare variables as time series
> rem #Declare variables as time series
> rem rtc exr #bind the data
> DataTC #Check correlations
> correlations #Check correlations
> correlations impact impact impact impact impact