Replace deprecated "_phase" suffix in reproin heuristic with "part-phase" entity #770
Conversation
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #770 +/- ##
==========================================
+ Coverage 82.08% 82.15% +0.06%
==========================================
Files 42 42
Lines 4215 4230 +15
==========================================
+ Hits 3460 3475 +15
Misses 755 755 ☔ View full report in Codecov by Sentry. |
yeah, but complicates review a little due to diff now including changes I can ignore or not and need to decide when looking... Could you submit a separate PR with that typo fix and such a rename (so no functionality change) and then let's improve this one -- yet to grasp on all the changes, and as we do not have any nice testing for reproin heuristic, I wonder how we could add one :-/ I would really be shy to accept without some way to test. Locally I could do using smth like https://github.com/nipy/heudiconv/blob/master/utils/test-compare-two-versions.sh but could we do better?! |
|
👍 I've opened #779 with the variable name change. |
| if "P" in curr_seqinfo.image_type: | ||
| series_info["part"] = "phase" | ||
| elif "M" in curr_seqinfo.image_type: | ||
| # if next one is phase from same scan, we should set part to mag |
There was a problem hiding this comment.
is it guaranteed to be the "next", could it be "prior"?
| "_SBRef" | ||
| ): | ||
| if "P" in curr_seqinfo.image_type: | ||
| series_info["part"] = "phase" |
There was a problem hiding this comment.
codecov says that this line is not covered, not yet sure how wide implication is ;-)
| and datatype_suffix | ||
| and datatype_suffix.startswith("scout") | ||
| and filename_suffix_parts[-1] | ||
| and filename_suffix_parts[-1].startswith("scout") |
There was a problem hiding this comment.
what does this have to do with phase suffix? nothing in short commit messages hints on possible relation to this change... why is it?
|
|
||
| next_series_description = next_seqinfo.series_description | ||
| next_dcm_image_iod_spec = next_seqinfo.image_type[2] | ||
|
|
There was a problem hiding this comment.
this duplicates that _replace hack 2 more times, causes me duplication-allergy...
Since we are to do that anyways, do it once then -- just in a loop before this loop apply the replacements to all first.
Then avoid that i_acq - do following 'zip'ing, as chatgpt suggested
my_list = [1, 2, 3, 4, 5]
for previous, current, next in zip([None] + my_list[:-1], my_list, my_list[1:] + [None]):
print(f"Previous: {previous}, Current: {current}, Next: {next}")
which would avoid all the if's and else's etc and make code easier to read IMHO.
|
@tsalo -- ping on this. Would you happen to get time to address comments? |
|
Sorry, I've been really busy. I might have time in the next couple of weeks though. |
Closes #769. This is working on DICOMs associated with https://openneuro.org/datasets/ds005250.
Changes proposed:
infotodict.RenameMoved to Rename s variable to curr_seqinfo in reproin heuristic #779.svariable tocurr_seqinfofor improved clarity.series_descriptionand is a phase scan. If it is, set "part" to "mag".