Match Stats

MATCH_STATS

See the source code.

This test verifies that region statistics are consistent between spatially and spectrally matched FITS and CASA images using various region types (rectangle, ellipse, polygon).

1. Frontend sends: OPEN_FILE (OpenFile) for two files

   file = "HD163296_CO_2_1.fits", file_id = 100
   file = "HD163296_CO_2_1.image", file_id = 101
   

2. Backend returns: OPEN_FILE_ACK for each file

3. Frontend sends: SET_REGION (SetRegion) for 4 regions on file 100

   Region 1 (Rectangle):

   region_id = 1
   region_type = RECTANGLE
   control_points = [{x: 250, y: 200}, {x: 300, y: 300}]
   rotation = 0
   

   Region 2 (Rotated Rectangle):

   region_id = 2
   region_type = RECTANGLE
   control_points = [{x: 350, y: 350}, {x: 100, y: 150}]
   rotation = 25
   

   Region 3 (Ellipse):

   region_id = 3
   region_type = ELLIPSE
   control_points = [{x: 150, y: 150}, {x: 60, y: 100}]
   rotation = 25
   

   Region 4 (Polygon):

   region_id = 4
   region_type = POLYGON
   control_points = [{x: 100, y: 150}, {x: 400, y: 400}, {x: 300, y: 30}]
   

4. Frontend sends: SET_STATS_REQUIREMENTS (SetStatsRequirements) for each region on both files

   file_id = <100 or 101>
   region_id = <1-4>
   stats_types = [NumPixels, Sum, FluxDensity, Mean, RMS, Sigma, SumSq, Min, Max]
   coordinate = "z"
   

5. Backend returns: REGION_STATS_DATA (RegionStatsData) for each region/file combination

Check 1: the REGION_STATS_DATA should satisfy:

• Each response has statistics.length > 0

• For each of the 4 regions, and for each of the 9 stats types: the value for file 100 (FITS) should be equal to the value for file 101 (CASA) within precision of 4 digits

• If either value is NaN, both must be NaN

MATCH_STATS_BORDERLINE

See the source code.

This test verifies region statistics consistency between FITS and CASA images when regions are positioned at or near image boundaries.

1. Frontend sends: OPEN_FILE (OpenFile) for two files

   file = "HD163296_CO_2_1.fits", file_id = 0
   file = "HD163296_CO_2_1.image", file_id = 1
   

2. Frontend sends: SET_REGION (SetRegion) for 4 boundary regions on file 0

   Region 1 (Rectangle at boundary):

   region_type = RECTANGLE
   control_points = [{x: 200, y: 150}, {x: 300, y: 300}]
   rotation = 0
   

   Region 2 (Rotated Rectangle at boundary):

   region_type = RECTANGLE
   control_points = [{x: 354, y: 343}, {x: 100, y: 150}]
   rotation = 25
   

   Region 3 (Ellipse at boundary):

   region_type = ELLIPSE
   control_points = [{x: 110, y: 250}, {x: 100, y: 200}]
   rotation = 45
   

   Region 4 (Polygon at boundary):

   region_type = POLYGON
   control_points = [{x: 100, y: 150}, {x: 429, y: 429}, {x: 300, y: 30}]
   

3. Frontend sends: SET_STATS_REQUIREMENTS for each region on both files

   stats_types = [NumPixels, Sum, FluxDensity, Mean, RMS, Sigma, SumSq, Min, Max]
   coordinate = "z"
   

Check 1: the statistics for file 0 (FITS) and file 1 (CASA) should be equal within precision of 4 digits for all regions and stats types

MATCH_STATS_WIDE

See the source code.

This test verifies region statistics consistency between FITS and CASA wide-field images using various region types.

1. Frontend sends: OPEN_FILE (OpenFile) for two files

   file = "casa_wideField.fits", file_id = 0
   file = "casa_wideField.image", file_id = 1
   

2. Frontend sends: SET_REGION (SetRegion) for 4 regions on file 0

   Region 1 (Rectangle):

   region_type = RECTANGLE
   control_points = [{x: 400, y: 1800}, {x: 800, y: 800}]
   rotation = 0
   

   Region 2 (Rotated Rectangle):

   region_type = RECTANGLE
   control_points = [{x: 1800, y: 800}, {x: 1000, y: 1000}]
   rotation = 45
   

   Region 3 (Ellipse):

   region_type = ELLIPSE
   control_points = [{x: 1800, y: 1300}, {x: 230, y: 300}]
   rotation = 22
   

   Region 4 (Polygon):

   region_type = POLYGON
   control_points = [{x: 3300, y: 1300}, {x: 3400, y: 120}, {x: 2200, y: 100}]
   

3. Frontend sends: SET_STATS_REQUIREMENTS for each region on both files

Check 1: the statistics for file 0 (FITS) and file 1 (CASA) should be equal within precision of 4 digits for all regions and stats types

MATCH_STATS_WIDE_BORDERLINE

See the source code.

This test verifies region statistics consistency between FITS and CASA wide-field images when regions cross image boundaries.

1. Frontend sends: OPEN_FILE (OpenFile) for two files

   file = "casa_wideField.fits", file_id = 0
   file = "casa_wideField.image", file_id = 1
   

2. Frontend sends: SET_REGION (SetRegion) for 4 boundary-crossing regions on file 0

   Region 1 (Rectangle near origin):

   region_type = RECTANGLE
   control_points = [{x: 300, y: 200}, {x: 400, y: 400}]
   rotation = 0
   

   Region 2 (Rotated Rectangle at edge):

   region_type = RECTANGLE
   control_points = [{x: 1800, y: 659.2}, {x: 1000, y: 1000}]
   rotation = 25
   

   Region 3 (Large Ellipse crossing boundary):

   region_type = ELLIPSE
   control_points = [{x: 1800, y: 1200}, {x: 800, y: 2000}]
   rotation = 330
   

   Region 4 (Polygon with negative coordinate):

   region_type = POLYGON
   control_points = [{x: 3500, y: 1300}, {x: 3599, y: -1.5}, {x: 2200, y: 100}]
   

3. Frontend sends: SET_STATS_REQUIREMENTS for each region on both files

Check 1: the statistics for file 0 (FITS) and file 1 (CASA) should satisfy:

• For NumPixels: the absolute difference should be <= 2 (allowing slight pixel count differences due to wide-field projections at boundaries)

• For all other stats types (Sum, FluxDensity, Mean, RMS, Sigma, SumSq, Min, Max): values should be equal within precision of 4 digits