PALS Python 0.3.0, Unit Convention for Quad#1357
Conversation
ax3l
added
bug
|
Oh, probably needs to update the analysis in |
|
@ax3l In the ImpactX parsing lines, we just need to check whether PALS input |
Merging this PR will degrade performance by 5.31%
|
| Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|
| ❌ | test_Drift[spin] |
457.4 µs | 483 µs | -5.31% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing cemitch99:update_pals_support (cf1c31a) with development (82163d6)
9a1e4af to
fa61eba
Compare
ax3l
changed the title
ax3l
changed the title
| ) | ||
|
|
||
|
|
||
| def flatten_pals(pals_data, registry=None): |
There was a problem hiding this comment.
flatten_pals could be a helper routine in pals-python, for the next release of it.
I am surprised that we still have placeholders at this point in the logic, we probably should replace placeholders / do lattice expansion already in pals.load() -- to check if as an optional argument (i.e. lattice_expansion=True). To check lattice and branch expansion.
There was a problem hiding this comment.
For the next release (not this PR): pals-project/pals-python#70
This comment was marked as resolved.
This comment was marked as resolved.
Update dependencies, not yet API changes.
New schema used in PALS.
- flatten PALS structure - address beaking API change in multipole/quad
Mirror MAD-X Logic Location
2 participants
Update PALS-Python from 0.2.0 to 0.3.0
Fix unit convention for PALS Quad support
PALS support requires additional implementation of translation on the ImpactX side.
For the PALS Quad, the option of specifying gradient in two unit systems, as
Bn1orKn1, needs to be supported. This PR fixes the existing choice of units.