Anonymization by Anonymouse.org ~ Adverts
Anonymouse better ad-free, faster and with encryption?
Enter a numeric expression:
I wrote a function called deconstruct for How JavaScript Works.
It breaks a binary floating point number down to its fundamental parts.
You can use those parts to precisely compute the actual numeric value.
If the number includes a decimal point or is very large, it is likely that its actual value is different than its literal value.
This difference is called an error. This is caused by having a basis of 2.
That makes it impossible to exactly represent most decimal number.
This is only a problem because humans use decimal numbers.
DEC64 does not have this error because it has a basis of 10. Note however that no computer can exactly represent any irrational number.