Skip to content

Added options to control argsort algorithm #148

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
daducci merged 2 commits into from
Aug 13, 2025
Merged

Added options to control argsort algorithm #148

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ac8688c
Added options to control argsort algorithm
daducci Aug 13, 2025
d7ad95f
Set mergesort as default algo for argsort (as it was before)
daducci Aug 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@
- Lesion mask support using ScalarMap models.
- Debias option in `fit()` to remove bias from the estimated coefficients when using regularization (e.g. `COMMIT2`).
- Public `dictionary_info` attribute on `Evaluation` for reuse (e.g. lesion masks, blur settings, thread counts).
- Two options to change the `argsort` algorithm (`sortDictionaryWith` and `sortThreadsWith`)

### 🛠️Changed
- `VolumeFractions` model now is called `ScalarMap` for clarity.
Expand All @@ -29,7 +30,7 @@

## `v2.3.0`<br>_2024-07-04_
### ✨Added
- Added support for Windows (requires the `pthread-win32` library)
- Added support for Windows (requires the `pthread-win32` library)
- Precompiled wheels for Windows, MacOS, and Linux are now available on PyPI

### 🛠️Changed
Expand Down
18 changes: 14 additions & 4 deletions commit/core.pyx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -129,6 +129,9 @@ cdef class Evaluation :
self.set_config('doDemean', False)
self.set_config('doNormalizeMaps', False)

self.set_config('sortDictionaryWith', 'mergesort')
self.set_config('sortThreadsWith', 'mergesort')

ui.set_verbose( 'core', self.verbose )


Expand Down Expand Up @@ -445,9 +448,13 @@ cdef class Evaluation :
if self.niiDWI is None :
logger.error( 'Data not loaded; call "load_data()" first' )

if self.get_config('sortDictionaryWith') not in ['quicksort','mergesort','stable']:
logger.error('The option "sortDictionaryWith" should be one of ["quicksort", "mergesort","stable"]')

tic = time.time()
logger.subinfo('')
logger.info( 'Loading the data structure' )
logger.debug( f'sortDictionaryWith = {self.get_config("sortDictionaryWith")}' )
self.DICTIONARY = {}

if path is not None:
Expand Down Expand Up @@ -499,7 +506,7 @@ cdef class Evaluation :
self.DICTIONARY['IC']['nF'] = self.DICTIONARY['TRK']['norm'].size

# reorder the segments based on the "v" field
idx = np.argsort( self.DICTIONARY['IC']['v'], kind='mergesort' )
idx = np.argsort( self.DICTIONARY['IC']['v'], kind=self.get_config('sortDictionaryWith') )
self.DICTIONARY['IC']['v'] = self.DICTIONARY['IC']['v'][ idx ]
self.DICTIONARY['IC']['o'] = self.DICTIONARY['IC']['o'][ idx ]
self.DICTIONARY['IC']['fiber'] = self.DICTIONARY['IC']['fiber'][ idx ]
Expand Down Expand Up @@ -530,7 +537,7 @@ cdef class Evaluation :
self.DICTIONARY['EC']['nE'] = self.DICTIONARY['EC']['v'].size

# reorder the segments based on the "v" field
idx = np.argsort( self.DICTIONARY['EC']['v'], kind='mergesort' )
idx = np.argsort( self.DICTIONARY['EC']['v'], kind=self.get_config('sortDictionaryWith') )
self.DICTIONARY['EC']['v'] = self.DICTIONARY['EC']['v'][ idx ]
self.DICTIONARY['EC']['o'] = self.DICTIONARY['EC']['o'][ idx ]
del idx
Expand All @@ -550,7 +557,7 @@ cdef class Evaluation :
self.DICTIONARY['nV'] = self.DICTIONARY['MASK'].sum()

# reorder the segments based on the "v" field
idx = np.argsort( self.DICTIONARY['ISO']['v'], kind='mergesort' )
idx = np.argsort( self.DICTIONARY['ISO']['v'], kind=self.get_config('sortDictionaryWith') )
self.DICTIONARY['ISO']['v'] = self.DICTIONARY['ISO']['v'][ idx ]
del idx

Expand Down Expand Up @@ -595,6 +602,8 @@ cdef class Evaluation :
logger.error( 'Dictionary not loaded; call "load_dictionary()" first' )
if self.KERNELS is None :
logger.error( 'Response functions not generated; call "generate_kernels()" and "load_kernels()" first' )
if self.get_config('sortThreadsWith') not in ['quicksort','mergesort','stable']:
logger.error('The option "sortThreadsWith" should be one of ["quicksort", "mergesort","stable"]')

self.THREADS = {}
self.THREADS['n'] = n
Expand All @@ -608,6 +617,7 @@ cdef class Evaluation :
logger.subinfo('')
logger.info( 'Distributing workload to different threads' )
logger.subinfo(f'Number of threads: {n}', indent_char='*', indent_lvl=1 )
logger.debug( f'sortThreadsWith = {self.get_config("sortThreadsWith")}' )

# Distribute load for the computation of A*x product
log_list = []
Expand Down Expand Up @@ -664,7 +674,7 @@ cdef class Evaluation :
if self.DICTIONARY['IC']['n'] > 0 :
self.THREADS['ICt'] = np.full( self.DICTIONARY['IC']['n'], n-1, dtype=np.uint8 )
if n > 1 :
idx = np.argsort( self.DICTIONARY['IC']['fiber'], kind='mergesort' )
idx = np.argsort( self.DICTIONARY['IC']['fiber'], kind=self.get_config('sortThreadsWith') )
C = np.bincount( self.DICTIONARY['IC']['fiber'] )
t = tot = i1 = i2 = 0
N = np.floor(self.DICTIONARY['IC']['n']/n)
Expand Down