I have a distribution of temperatures available to me.
Currently, my algorithm is as follows:
if (temperature <= 25th percentile of distribution) or (elapsed time >= timeout):
do something
The above logic checks to see if the current temperature is within 25th percentile of the available distribution of temperatures or if the total elapsed time has exceeded a fixed value. If either condition is satisfied, it runs some logic.
One way to make this algorithm more 'flexible' would be to linearly relax the constraint based on the percentage of time that has elapsed:
percentage = (elapsed time * 100) / timeout
new constraint = min(25 + percentage, 100)
Is there a technical name for these kind of functions?
Is there a formal method for relaxing the constraint (25th percentile) based on the amount of time that has elapsed?
Could you give me examples of a few other functions and describe how they would behave?