chenlin/Mujoco_WASM
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add more details to reproducibility section.

Browse Source 
PiperOrigin-RevId: 624186625
Change-Id: Ibceace5b2b3d84c9cf0a179adfb6638450f9cb9f
This commit is contained in:
Kevin Zakka
2024-04-12 08:26:01 -07:00
committed by Copybara-Service
parent 9463a8816d
commit bcee4e9954
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
doc/computation/index.rst
+4 -4
View File
@@ -1663,10 +1663,10 @@ important caveats:
systems with contact. Contact events have high `Lyapunov exponents systems with contact. Contact events have high `Lyapunov exponents
<https://en.wikipedia.org/wiki/Lyapunov_exponent>`__; this is a property of any rigid-body simulator (and indeed of <https://en.wikipedia.org/wiki/Lyapunov_exponent>`__; this is a property of any rigid-body simulator (and indeed of
`real-world physics <https://en.wikipedia.org/wiki/Roulette>`__) and is not MuJoCo-specific. `real-world physics <https://en.wikipedia.org/wiki/Roulette>`__) and is not MuJoCo-specific.
- Exact reproducibillity is only guaranteed within a **single version**. Small numerical differences are quite common - Exact reproducibility is only guaranteed within a **single version**, on the **same architecture**. Small numerical
between versioned releases, for example due to code optimizations. This means that when saving an initial state and an differences are quite common between versioned releases, for example due to code optimizations. This means that when
open-loop control sequence, the resulting rolled-out trajectory will be identical within the same version but will saving an initial state and an open-loop control sequence, the resulting rolled-out trajectory will be identical
likely be different between MuJoCo versions. within the same version, but will likely be different between MuJoCo versions or different operating systems.
.. _derivatives: .. _derivatives: