Review sampling and reweighting algorithms
As the example in #608 shows, there are situations where the integration step appears ok, but doesn't catch the true maxima of the integrand. In particular, if there are lots of channels, the distribution of the sampling points (stratified vs. importance, weighting of channels) might be sub-optimal.
At some point, we should return to this basic issue, taking typical processes as examples.
Change History (4)
| Milestone: |
v2.2.3 →
v2.4.0
|
| Resolution: |
→ wontfix
|
| Status: |
new →
closed
|
| Milestone: |
v2.4.0 →
v2.5.0
|
| Milestone: |
v2.5.0 →
v2.6.0
|
Milestone renamed