Verify XXXX Client 3 days Look Back Window

I have the sample data uploaded. NOTICE: The data was different from the actual data. For the survival analysis, I use the lifelines package. Besides, I use Pandas, Numpy, maplotlib, datetime, warnings, pylab.

You can install lifelines using

pip install lifelines

I implemented Bootstrapping AB testing because of the unbalance data.

df.sample(frac=0.5, replace=True)

And draw the Kernel Density Estimation(kde) of the two groups difference.

• Minimum Detectable effer: Unknown
  So I will draw the conclusion based on the 95% confidence interval.
  

Please refer the chart below on how I draw conclusion on significant or not.