Anonymous user
Talk:Perlin noise: Difference between revisions
m
→exactitudeness: added comments about exact value(s) using decimal floating point.
m (→exactitudeness) |
m (→exactitudeness: added comments about exact value(s) using decimal floating point.) |
||
Line 28:
<br>
''... show that the Perlin noise ... of a point in 3D-space with coordinates 3.14, 42, 7 is 0.13691995878400012.''
<br>Is that number cast in stone, or is it a result of particular type of binary floating point arithmetic?
Line 49:
: That result was achieved using numbers represented in the 64 bit IEEE-754 format. --[[User:Rdm|Rdm]] ([[User talk:Rdm|talk]]) 13:26, 7 September 2015 (UTC)
:: Yes, but the ''exact'' value is 0.136919958784 when computed via (true) ''floating point decimal'' (with no rounding). This should've been noted in the task's preamble. For other 3D points, more decimal precision may be needed. For the 3D point specified in this task, only 12 decimal digits are needed to calculate the exact result. -- [[User:Gerard Schildberger|Gerard Schildberger]] ([[User talk:Gerard Schildberger|talk]]) 20:47, 16 June 2016 (UTC)
| |||