Skip to content

RADIANCE variability is breaking current pytests #593

Description

@cdeline

rtrace.exe has slight differences from one RADIANCE release to another. Specifically there was a change in behavior between two releases between 9e3e6287 on april 19 2022 and release b09c6327 on april 21 2022.

Prior to the April 21 release, if the rays traced by rtrace using ambient super-sampling encountered a perfectly black region (zero luminance/reflectance), a bug caused the software to generate and return NaN values. This artificially inflated returned irradiance values. The variability in this result was specifically encountered in the bifacial_radiance pytest test_1axis_gencumSky() .

The solution here is: specifically make the test_1axis_gencumSky() pytest use the 'accuracy=high' option in analysis1axis, which will reduce the variability in this scan, and update passing values to 190 +/-15 instead of 210 +/- 30. Also, propose that we make 'accuracy=high' the default option for all analysis and analysis1axis calls. Also, in our CI test routine, we should load a more updated version of RADIANCE specifically > 6.0 which has the rtrace bugfix.

Metadata

Metadata

Assignees

Labels

v0_6_0Target for v0.6.0 release

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions