For each type of function there is a different error calculation
for let's say x = tsin(sy) the error would be
Δx = (Δy)tscos(sy) where t and s are some arbitrary constants
and something like
x = tzy were t is an arbitrary constant
then
(Δx/x)2 = (Δz/z)2 + (Δy/y)2 + 2(Δzy)2/zy
where 2(Δzy)2/zy is the covariance factor which i doubt you need to include.
so Δx = x√(all error added together and squared individually)
so
error calculations are a pain in the but in upper level physics studies but they are a necessity.
I'll give you an different example in case it doesn't make a lot of sense
Suppose that the area of a rectangle A=LW is to be determined from the following measurements of lengths of two sides:
L = 22.1 ± 0.1cm W= 7.3 ± 0.1cm
The relative contribution of ΔAL to the error in L will be
ΔAL/A = ΔL/L = 0.1/22.1 = 0.005
and the corresponding contribution of ΔAW will be
ΔAW/A = ΔW/W = 0.1/7.3 = 0.014
Thus ΔA will equal
ΔA = A√( 0.0142 + 0.0052)
ΔA = 0.015A
Last edited: