Sample N times from one metric, and perform operations on them.
Ozzie Gooen
It would be convenient to sample multiple times from one metric. For instance,
salesperson_average_monthly_revenue = 3K to 10K
total_monthy_revenue = sum(sample(salesperson_average_monthly_revenue,5)
Not sure if there's a better possible notation for this. One issue here is that it would only work for a narrow set of functions (sum, product.) Sample may also not be the best word here.
Old github feature: https://github.com/getguesstimate/guesstimate-app/issues/240
Alexander W. Janssen
What would be related: Assume I have a Poisson-distributed customer arrival rate with a mean of 5 per hour, and the average (normal-distributed) customer serving time is 5 minutes. How many minutes of work do I have to spend in each sample?
An example in R would be:
> customers_per_hour <- rpois(10, 5) # 10 samples, mean=5
> customers_per_hour
[1] 5 4 3 6 8 5 3 5 6 4
> sapply(customers_per_hour, function(x) sum(rnorm(x, 5)))
[1] 27.32224 23.75343 16.46017 28.59986 38.02239 26.07986 13.15253 25.70241
[9] 33.46726 21.67366
Ozzie Gooen
From one user:
"I would probably want to do: sum of samples, product of samples, get percentile from a cell's distribution (e.g. median), mean."
This was to understand home value over time.