Return early when no slicing needed

[andy-sweet]

Description

This fixes the slicing performance issues detected in #5194. These were caused by the refactor in #5003, which removed the case where we return early in Layer._slice_indices when no slicing is actually needed.

Type of change

• Bug-fix (non-breaking change which fixes an issue)

References

Closes #5194

How has this been tested?

• all existing tests pass with my change

I ran some of the slicing benchmarks affected by this in #5194 at a few different commits.

First, at the commit on main corresponding to #5003 (2566980).

(napari-dev) ➜  napari git:(256698021) asv run --bench "Image2DSuite.time_set_view_slice" --python=same
· Discovering benchmarks
· Running 2 total benchmarks (1 commits * 1 environments * 2 benchmarks)
[  0.00%] ·· Benchmarking existing-py_Users_asweet_software_miniconda3_envs_napari-dev_bin_python
[ 25.00%] ··· Running (benchmark_image_layer.Image2DSuite.time_set_view_slice--)..
[ 75.00%] ··· benchmark_image_layer.Image2DSuite.time_set_view_slice                                                                                                                                                                                                      ok
[ 75.00%] ··· ======== ==========
               param1            
              -------- ----------
                 16     717±20μs 
                 32     714±10μs 
                 64     752±10μs 
                128     755±70μs 
                256     768±40μs 
                512     736±9μs  
                1024    743±10μs 
                2048    762±30μs 
                4096    720±30μs 
              ======== ==========

[100.00%] ··· benchmark_qt_slicing.AsyncImage2DSuite.time_set_view_slice                                                                                                                                                                                                  ok
[100.00%] ··· ======== ============ ======================
              --                      param2              
              -------- -----------------------------------
               param1   skin_data    jrc_hela-2 (scale 3) 
              ======== ============ ======================
                0.0     1.49±0.1ms        4.78±0.1ms      
                0.05     319±2ms           5.01±0s        
                0.1      616±2ms           9.78±0s        
              ======== ============ ======================

Then at the previous commit (e6ab7cd) to show the regression.

(napari-dev) ➜  napari git:(e6ab7cda8) asv run --bench "Image2DSuite.time_set_view_slice" --python=same
· Discovering benchmarks
· Running 2 total benchmarks (1 commits * 1 environments * 2 benchmarks)
[  0.00%] ·· Benchmarking existing-py_Users_asweet_software_miniconda3_envs_napari-dev_bin_python
[ 25.00%] ··· Running (benchmark_image_layer.Image2DSuite.time_set_view_slice--)..
[ 75.00%] ··· benchmark_image_layer.Image2DSuite.time_set_view_slice                                                                                                                                                                                                      ok
[ 75.00%] ··· ======== ===========
               param1             
              -------- -----------
                 16     92.1±30μs 
                 32      66.2±2μs 
                 64      66.3±3μs 
                128      64.3±5μs 
                256      66.3±2μs 
                512      65.3±2μs 
                1024     66.1±4μs 
                2048     72.0±6μs 
                4096     66.2±3μs 
              ======== ===========

[100.00%] ··· benchmark_qt_slicing.AsyncImage2DSuite.time_set_view_slice                                                                                                                                                                                                  ok
[100.00%] ··· ======== =========== ======================
              --                     param2              
              -------- ----------------------------------
               param1   skin_data   jrc_hela-2 (scale 3) 
              ======== =========== ======================
                0.0      736±60μs        4.85±0.3ms      
                0.05    320±0.9ms        5.06±0.01s      
                0.1     614±0.7ms        9.80±0.02s      
              ======== =========== ======================

Then with this PR, to show the fix.

(napari-dev) ➜  napari git:(fix-5194-early-return-slice) asv run --bench "Image2DSuite.time_set_view_slice" --python=same
· Discovering benchmarks
· Running 2 total benchmarks (1 commits * 1 environments * 2 benchmarks)
[  0.00%] ·· Benchmarking existing-py_Users_asweet_software_miniconda3_envs_napari-dev_bin_python
[ 25.00%] ··· Running (benchmark_image_layer.Image2DSuite.time_set_view_slice--)..
[ 75.00%] ··· benchmark_image_layer.Image2DSuite.time_set_view_slice                                                                                                                                                                                                        ok
[ 75.00%] ··· ======== ============
               param1              
              -------- ------------
                 16      68.9±9μs  
                 32     67.6±0.7μs 
                 64      69.3±6μs  
                128      64.5±2μs  
                256      67.8±1μs  
                512      64.8±1μs  
                1024     67.7±2μs  
                2048     67.5±2μs  
                4096     70.3±3μs  
              ======== ============

[100.00%] ··· benchmark_qt_slicing.AsyncImage2DSuite.time_set_view_slice                                                                                                                                                                                                    ok
[100.00%] ··· ======== =========== ======================
              --                     param2              
              -------- ----------------------------------
               param1   skin_data   jrc_hela-2 (scale 3) 
              ======== =========== ======================
                0.0      700±8μs         5.15±0.8ms      
                0.05     320±1ms          5.03±0s        
                0.1     614±0.8ms        9.85±0.01s      
              ======== =========== ======================

Final checklist:

• My PR is the minimum possible work for the desired functionality
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works
• If I included new strings, I have used trans. to make them localizable.
  For more information see our translations guide.

[codecov]

Codecov Report

Merging #5239 (13efd57) into main (fb81ff6) will increase coverage by 0.04%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #5239      +/-   ##
==========================================
+ Coverage   88.74%   88.79%   +0.04%     
==========================================
  Files         574      574              
  Lines       48954    48956       +2     
==========================================
+ Hits        43446    43472      +26     
+ Misses       5508     5484      -24     

Impacted Files	Coverage Δ
napari/layers/base/base.py	91.49% <100.00%> (+0.02%)	⬆️
napari/components/experimental/chunk/_pool.py	85.71% <0.00%> (-7.94%)	⬇️
napari/_qt/qt_main_window.py	75.41% <0.00%> (+0.18%)	⬆️
napari/layers/image/image.py	95.89% <0.00%> (+0.24%)	⬆️
napari/utils/theme.py	89.79% <0.00%> (+0.68%)	⬆️
napari/_qt/qt_event_loop.py	82.31% <0.00%> (+0.68%)	⬆️
napari/utils/interactions.py	74.41% <0.00%> (+2.32%)	⬆️
napari/layers/image/_image_slice.py	100.00% <0.00%> (+2.38%)	⬆️
napari/_qt/widgets/qt_color_swatch.py	74.79% <0.00%> (+3.25%)	⬆️
napari/utils/info.py	83.33% <0.00%> (+4.44%)	⬆️
... and 2 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[andy-sweet]

Thanks for the reviews @jni and @Czaki. I will merge this after 24 hours.

[kcpevey]

Nice fix! :)