On 2014-01-03, Dave Angel <da...@davea.name> wrote:
> On Thu, 2 Jan 2014 16:23:22 +0000 (UTC), Grant Edwards
><invalid@invalid.invalid> wrote:
>> AFAIK, that's irrelevent. time.time() returns a float. On all the
>> CPython implementations I know of, that is a 64-bit IEEE format,
>> which provides 16 decimal digits of precision regardless of the
>> granularity of the system time value. At this point in time, that
>> means 10 digits left of the decimal point and 6 to the right.
>
> Correction: no more than about 6 to the right. You can certainly get
> less, from an os with a smaller resolution.
time.time() returns a Python float. A Python float will have 16 digits
of precision. Perhaps the OS always sets some of those digits to 0 (or
even random values), but they're still there. Perhaps the accuracy or
granularity of the values returned is problematic on some OSes, but
the precision of the value doesn't change: there's no way he's "only
getting 2 decimal places" from time.time() unless (as you mention
below) he's printing them using a method that truncates/rounds.
> Or you can lose some of what you do get by printing in a sub-optimal
> way.
Yes, depending on how you print the value, you can hide some of the
digits. But, there's no way for time.time() to return a value with
less than ~16 decimal digits of precicsion.
--
Grant Edwards grant.b.edwards Yow! ! Everybody out of
at the GENETIC POOL!
gmail.com
--
https://mail.python.org/mailman/listinfo/python-list
