Algorithm to smooth numbers with variable input time -

Algorithm to smooth numbers with variable input time -

i have app accepts integers @ variable rate every .25 2 seconds.

i'd output info in smoothed format 3, 5 or 7 seconds depending on user input.

if info came in @ same rate, let's every .25 seconds, easy. variable rate confuses me.

data might come in this: time - data 0.25 - 100 0.50 - 102 1.00 - 110 1.25 - 108 2.25 - 107 2.50 - 102 ect...

i'd display 3 sec rolling average every .25 seconds on display.

the simplest form of doing set each item array time stamp.

array.push([0.25, 100]) array.push([0.50, 102]) array.push([1.00, 110]) array.push([1.25, 108])

ect...

then every .25 seconds read through array, front, until got time less now() - rollingaveragetime. sum , display it. .shift() origin of array.

that seems not efficient though. wondering if had improve way this.

why don't save timestamp of starting value , accumulate values , number of samples until timestamp >= startingtime + rollingaveragetime , split accumulator number of samples taken?

edit: if want preserve number of samples, can way:

take accumulator, , each input value sum , store value , timestamp in shift register; @ every cycle, have compare latest sample's timestamp oldest timestamp in shift register plus smoothing time; if it's equal or more, subtract oldest saved value accumulator, delete entry shift register , output accumulator, divided smoothing time. if iterate obtain rolling average (i think) to the lowest degree amount of computation each cycle:

a sum (to increment accumulator) a sum , subtraction (to compare timestamp) a subtraction (from accumulator) a partition (to calculate average, done in smart way can shift right)

for total of 4 algebric sums , partition (or shift)

edit:

for taking business relationship time lastly sample weighting factor, can split value ratio between time , averaging time, , obtain weighted average, without having split accumulator.

i added part because doesn't add together computational load, can implement quite easy if want to.

algorithm