[kgm]

For another breakdown of the details of floating-point precision, I recommend qntm's "0.1 + 0.2 returns 0.30000000000000004":

https://qntm.org/notpointthree