Sadly our IT has not installed us the Predictive Tools at the moment but i wanted to see if i can do the same in Python. I have to say my solution is done without spending too much work so it could be done much better. Hopefully when we get the Tools i can compare both ways.
I missed the hints in the description so I could have handled this better, but still...
#set threshold for 1 person that can reasonably be expected to be >.9 or <.5
z <- qnorm(0.9995)
#generate normally distributed data where 99.9% of samples are between .5 and .9
Score <- rnorm(1000, mean=0.7, sd=(0.2/z))
#reset values below 0.5 to 0.5 and above 0.9 to 0.9
Score <- ifelse(Score < 0.5, 0.5, Score)
Score <- ifelse(Score > 0.9, 0.9, Score)
#ensure the dataset is 1 column with 1000 rows (from 1 vector to 1 dataframe)
Score <- data.frame(Score)
#write to Alteryx