Questions tagged [causalimpact]
CausalImpact R package for estimating the effect of an intervention on a time series.
91
questions
0
votes
0answers
14 views
Selecting best Causal Impact model from available set of different models [closed]
I am working on figuring out methodology to find best Causal Impact model from available set of different models generated by Causal Impact Library. Currently, I am exploring cross-validation approach ...
0
votes
0answers
4 views
Causal Impact R - Ending the intervention period before the end of the time series [closed]
so I am using Causal impact, and my intervention period is continuous over a week, then ends and the data should pretty much go back to normal. (for example if i was looking at number of customers in ...
0
votes
0answers
20 views
How to choose control groups (and how many) to use in CausalImpact?
I'm using the CausalImpact package and noticed that the results are very sensitive to the choice of control groups being fed. My main questions are:
Is there a recommended approach for how to choose ...
0
votes
0answers
38 views
Backtesting for Causal Impact and multiple testing correction
I plan to run a CausalImpact model to estimate the effect of an intervention on my metric of interest. In order to estimate the real false positive rate and the real power of the experiment given the ...
0
votes
0answers
29 views
Causal impact predictions : How to account for holidays in post period?
I'm trying to use the causal impact package to test the effects of a product release. My data is at a day level and I'm using a nseasons=7 paramater, which produces ...
0
votes
0answers
20 views
CausalImpact - Estimated coefficients of predictor variables
I am using the R package CausalImpact and in my model I am including several predictor variables.
I know that I can plot the posterior probability of each predictor being included in the model using
&...
0
votes
1answer
33 views
CausalImpact for a single individual time series of physiological data
I am using CausalImpact for the first time and I am not completely sure whether it would be statistically correct for my case. We wanted to test if the change of enclosure did significantly impact the ...
0
votes
0answers
31 views
If correcting p values for multiple comparisons in backtesting, do I need to correct the actual test's p value too?
My question relates to p-value correction in the context of causal inference using the causal impact method developed by Brodersen et al. for Google [link].
In trying to validate the accuracy and ...
0
votes
0answers
27 views
Picking a suitable performance metric when comparing the same model but using different sets of training data (Causal inference model)
I am comparing the same models prediction accuracy (Causal Impact) using different control variables as predictors and looking for a metric to decide which set of controls to use. Reading into AIC and ...
0
votes
0answers
26 views
Customized model in CausalImpact library whitout exogenous variable (python)
I am interested in using the CausalImpact library that uses Bayesian structural time series models. However I can't find a way to define my own model without incorporating an exogenous variable. When ...
0
votes
1answer
115 views
Generalized difference-in-differences where all units are treated, but at a different intensity
Is it possible to run a generalized difference-in-differences analysis where all units are treated, but at a different intensity?
In my situation, a policy was introduced in 5 countries, designed to ...
2
votes
1answer
125 views
Inferring causal effects on a time series from a forecast
Part of my job is measuring the effect of marketing interventions using experiments when possible, or estimating their effect when it's not possible to experiment.
I know relatively little about ...
0
votes
1answer
108 views
difference-in-difference: Dynamic treatment group/timing
I want to use difference-in-difference (DiD) to estimate a treatment effect. However, my problem is a little different from the standard DiD application in that:
The items in the treatment group may ...
2
votes
0answers
41 views
how to measure the causal impact of interventions which happen at different times on different time series?
My data consists of a bunch of time series of daily clicks on some merchandises on a website, a portion of which have an intervention (only 1 intervention per time series), others don't. The ...
2
votes
0answers
31 views
how to get the confidence interval of accumulative effect when using Difference-In-Difference
say i have a dataframe like below:
...
1
vote
1answer
57 views
Causal Inference on test scores
I administered a test and wanted to know if the exam scores were influenced by watching videos. The participants were randomly entered into 2 arms. I have one control arm that did not watch videos, ...
2
votes
1answer
54 views
Difference in difference with similar units over 2 periods of time
I have ESG (Environmental, Social & Governance) scores for 20 companies over a period of 10 years. In the fifth year a policy was introduced and I want to estimate the impact/effect of the policy ...
5
votes
2answers
148 views
causal impact estimation
say i have following causal model:
outcome variable: Y (e.g. sales)
treatment variable: T (e.g. price)
covariate variable: x2 (e.g. traffic)
unobserved variables: U (unobserved)
causal relation:
...
1
vote
0answers
60 views
how to run causalImpact on a time series with multiple interventions?
In CausalImpact package, one defines pre-period and post-period for a single intervention. However, in the scenarios where there are multiple interventions at irregular time intervals, is there a way ...
0
votes
0answers
28 views
What is a collider
The term "collider" is usually taken to mean a variable in a causal relationship that opens up spurious correlation -- usually between the outcome and the treatment variable (and hence would ...
0
votes
0answers
43 views
Modeling year-level seasonality in R-package CausalImpact for bayesian structural time-series
I have a time-series history of 2 years with spikes occcuring once a year (christmas time) and am trying to model it with a seasonal component tat catches this yearly spike.
However, no matter if I ...
0
votes
0answers
62 views
Why diff-in-diff over regression with control variable for measuring treatment intervention in time series?
Let's suppose we have a time series data set, with an intervention period for the treatment where the mean of the independent variable increases by 10. The goal is to measure the effect of the ...
0
votes
0answers
17 views
Isolating and Quantifying Price Change effects in Products
Let say, I changed prices in my products or changed rebates for quantity. Then, after a while I want to analyse what effect this particular measure caused on my revenues. Of course there are some ...
0
votes
1answer
65 views
CausalImpact plot upper and lower bounds
Below is a screenshot of causal impact result.
Upper and lower bounds of point wise results plot hovers around 10k and -10k. In summery statistics, sd shows 1879 with 95 CI of -4674 and 2736. Is the ...
0
votes
1answer
125 views
CausalImpact estimates a statistically significant impact for multiple intervention-free validation periods
While trying to validate the estimated impact by CausalImpact package I ran backtests on several dummy pre and post time-periods prior to the actual intervention but ended up with statistically ...
1
vote
1answer
162 views
Causal impact R - Is it possible to model multiple the pre and post periods? Indivudally for participants
I would like to use the causal impact algorithm, however, not in the context of marketing, but in medicine. The problem is that the intervention does not take place at the same time, but on an ...
2
votes
1answer
2k views
Staggered Diff-in-Diff: multiple treatments equation
I wonder if you can help me to figure out how to rewrite basic difference-in-difference equation (pictured) so that it takes into account the fact that treatment has occurred at different times for ...
1
vote
0answers
68 views
Combine Different Metrics to CausalImpact
I am currently trying to understand the Google package of CausalImpact, to be more concrete the python implementation.
My current state is that I have for example the sales of a car on different ...
0
votes
1answer
66 views
How to test the stability assumption when using the CausalImpact package in R?
In the case of CausalImpact it is assumed that the relationship between covariates and treated time series, as established during the pre-period, remains stable throughout the post-period.
Why is it ...
1
vote
0answers
118 views
How to use CausalImpact in R to construct an interrupted time series model with >2 periods of interest?
I would like to conduct an interrupted time series analysis on data with three time periods:
Pre intervention
During intervention
Post intervention
My outcome variable is a continuous biological ...
1
vote
0answers
95 views
Logistic Regression Coefficient Intrepretation - Impact On Churn
I am using logistic to see the relationship between churn and minute seen.
I get something like this
...
-1
votes
1answer
114 views
Using CausalImpact to estimate uplift attributed to a promotion for logged-in users
I'm working for an online travel agency, and we've been exploring the use of the causalImpact algorithm for the following problem:
Hotel owners that sell room-...
1
vote
1answer
160 views
These summary statements seem contradictory, or have I misunderstood something?
So here's the output summary I got when running an analysis.
This means that, although it may look as though the intervention has exerted a negative effect on the response variable when considering ...
0
votes
1answer
269 views
Causal Impact: how to show all data points in predicted series? [closed]
This might be a very basic code question, but thank you for your help. I have used the Causal Impact package to run a regression on a data set with 144 monthly observations, cutting at point 88. I ...
2
votes
1answer
275 views
Estimate the time series like an event was never happened
I have data from a website where a specific advertising campaign happened a couple of years ago. What I want to do is to estimate how the signups on that website would have been without that big ...
0
votes
1answer
92 views
Matching subjects with themselves when evaluating short term outcomes
I am considering a simple causal inference scenario; Let's say we want to examine the effect of paracetamol (treatment) on curing headache (outcome).
When performing matching, is it okay to match ...
1
vote
0answers
50 views
Ok to use PSM to create treatment groups and then plug into CausalImpact? [closed]
Is it ok to use propensity score matching to create treatment and control groups and then plug these two time series into CausalImpact to estimate your treatment effect? I might want to do this, for ...
1
vote
1answer
379 views
CausalImpact Conflicting Results - Posterior prob. v Confidence Interval
I'm getting conflicting results from a Causal Impact analysis I'm running. The 95% C.I. indicates that 0 is included and thus the impact is not significant; however the Posterior Probability is <....
1
vote
0answers
335 views
How should I specify seasonality in CausalImpact? [closed]
I am working with the Python implementation of Google's CausalImpact package. My data is at daily frequency (365 observations per year); however, to inspect the effect of intervention, my pre-period ...
0
votes
1answer
105 views
Using CausalImpact in R, observed data affected by event that could not affect priors
I am just starting out using the Causal Impact package in R. I am looking at observed data where I know there will be periods of extra traffic due to outside forces. However I cant find any unbiased ...
7
votes
3answers
278 views
How to calculate causal effects with repeated exogenous shocks over a time series
A rather frequent problem in causal inference is that we come across various shocks over time and try to measure their impact.
In the case of a single shock we can use bayesian methods to predict how ...
1
vote
0answers
151 views
Default CausalImpact Package
I am trying to gain an understanding of the CausalImpact package and the BSTS approach. Consider the following model:
$y_t = \mu_t + z_t + \epsilon_t$, where $\epsilon_t \sim N(0,s^2)$, and
$\mu_{t+1} ...
0
votes
1answer
213 views
How is the posterior tail-area probability calculated?
I am currently using the CausalImpact package for some research and in this context I need to know and be able to explain, how the posterior tail-area probability is calculated in order to reproduce ...
1
vote
0answers
59 views
inclusion of static variables as covariates [closed]
Is there a clear way to include static variables e.g. elevation as covariates in the CausalImpact pacakge. Any guidance would be greatly appreciated.
Anand
0
votes
1answer
133 views
Incorporating Fixed Effects with Causal Impact
I have observations on a dependent variable of interest, $Y_{ijt}$, where $t \in [1,2,...T]$ denotes the time, $i$ denotes a particular brand and $j$ a specific product item. At $t=k$, there was an ...
3
votes
1answer
439 views
Synthetic control and unobserved confounders
The synthetic control (cohort) method is a very promising approach to causal inference that has been used in a number of interesting studies. It's particularly useful in situations where data are only ...
0
votes
1answer
119 views
Pre.intervention and post.intervention should be contiguous in CausalImpact?
I am running a CausalImpact analysis on a time series and my pre.period goes from 01.01.15 to 30.03.15. I want my post period to be from 15.04.15 to 17.04.15.
Is it ok if I create a time series that ...
1
vote
0answers
37 views
Assesing impact for rolling enrollment data
I'm running an experiment where some of the users are shown a new widget when they become eligible (some conditions that are unrelated to the experiment itself). I have a control counterfactual group ...
1
vote
1answer
117 views
Is it possible to predict the pre-intervention period rather than the post-period using Google's CausalImpact function?
I want to use to Google's causal impact function to impute the effect of an intervetion. However, my data is structured as: pre-period=1991-1995, intervetion occurs, post-period=1996-2017. To clarify, ...
3
votes
2answers
1k views
How to choose Control groups for Causal Impact algorithm?
I'm running an experiment and want to use the Causal Impact function to assess how well it performs.
I have 10 different cities. I'm looking to find out what is the best method for choosing which ...
