[R] A very small p-value

[Ivan Krylov]

В Sat, 25 Oct 2025 22:37:57 +0200
Christophe Dutang <dutangc using gmail.com> пишет:

> But I was wondering why pt() returns something smaller than
> double.eps.

That's possible because the significant digits are stored separately
from the exponent of the number (which goes approximately from 10^-308
to 10^+308). Thus 5., followed by the rest of the digits, times 10^-19
can be distinguished from zero, despite the same number "vanishes" when
subtracted from 1.

If this answers the wrong "why", could you please rephrase the question?

-- 
Best regards,
Ivan