Load pulseq sequences faster and allow comparison #238
+320
−2
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This addresses #224. Currently a draft, because it hasn't been extensively tested on other pulseq versions.
The below snippet was used for comparison and also contains some further explanations.
An example sequence file is here bSSFP_FA30deg_TE10ms_TR20ms_2D_(69x64)_pulseq.seq.zip. This small sequence loads quite fast, but for e.g. 3D sequences with many excitations, the performance issues and benefits from this diff are greatly increased.
Proper unit tests should of course be required. What I did here was to implement
==
operator forsequences (and constituent structs) to compare structurally, rather than by ref, and base the
sequence equality on that. I assert that results of the existing implementation and this one are equal.
There may be an existing julia package to do that equality through a macro (StructEquality.jl),
rather than by hand, but I wanted to keep changes light.
There seemed to be a rather big refactor going on with KomaMRIFiles subpackage that threw my
environment off. As a result, this is based on an earlier version.