Percentiles don’t work: Analyzing the distribution of response times for web services
Adrian Cockcroft
JANUARY 29, 2023
For this demo on an old MacBook (2.7 > system.time(wait1 <- normalmixEM(waiting, mu=c(50,80), lambda=.5, > system.time(wait1 <- normalmixEM(waiting, mu=c(50,80), lambda=.5, The initial loglik value is higher but trends to the same result. That’s why an algorithm is needed to take out the guesswork.
Let's personalize your content