302 code bug spot analysis visualizations have nested window dressings

[bbean23]

Purpose

This fixes the nested window dressings issue by clearing window dressings before rendering on visualizations that are dependencies of later visualizations.

Fixes #302

Summary of changes

• Clears window dressings before rendering.
• Sets the figure size to eliminate extra whitespace.

Implementation notes

Correctness was verified with the following example code (from the upcoming CrossSection example looking straight at the sun):

def centroid_pixel_locator(operable: SpotAnalysisOperable) -> tuple[int, int]:
    """Returns the x/y pixel location of the centroid center"""
    moments = filter(lambda a: isinstance(a, MomentsAnnotation), operable.annotations)
    mom_annotation: MomentsAnnotation = list(moments)[-1]
    ret = mom_annotation.centroid.astuple()
    return (int(ret[0]), int(ret[1]))

def hotspot_pixel_locator(operable: SpotAnalysisOperable) -> tuple[int, int]:
    """Returns the x/y pixel location of the hotspot center"""
    hotspots = filter(lambda a: isinstance(a, HotspotAnnotation), operable.annotations)
    hs_annotation: HotspotAnnotation = list(hotspots)[-1]
    ret = hs_annotation.origin.astuple()
    return (int(ret[0]), int(ret[1]))

def rgb2gray(operable: SpotAnalysisOperable) -> np.ndarray:
    grayscale_image = cv.cvtColor(operable.primary_image.nparray, cv.COLOR_RGB2GRAY)
    return grayscale_image

image_processors = {
    "EchoEcho": EchoImageProcessor(),
    "Original": ConvolutionImageProcessor(diameter=1),
    "Rgb2Gray": CustomSimpleImageProcessor(rgb2gray),
    "PopStats": PopulationStatisticsImageProcessor(),
    "Centroid": MomentsImageProcessor(
        include_visualization=True, centroid_style=rcps.default(color=color.cyan(), markersize=20)
    ),
    "VFalseCl": ViewFalseColorImageProcessor(),
    "VCentOrg": ViewAnnotationsImageProcessor(base_image_selector='visualization'),
    "CropCent": CroppingImageProcessor(centered_location=centroid_pixel_locator, width_height=(1500, 1500)),
    "HotSpotS": HotspotImageProcessor(
        21, draw_debug_view=False, record_visualization=False, record_debug_view=False
    ),
    "VFalseC2": ViewFalseColorImageProcessor(),
    "VHotspot": ViewAnnotationsImageProcessor([HotspotAnnotation], base_image_selector='visualization'),
    "CropHots": CroppingImageProcessor(centered_location=hotspot_pixel_locator, width_height=(250, 250)),
    "Centrod2": MomentsImageProcessor(
        include_visualization=True, centroid_style=rcps.default(color=color.cyan(), markersize=20)
    ),
    "SpotSize": SpotWidthImageProcessor(spot_width_technique="fwhm"),
    "VFalseC3": ViewFalseColorImageProcessor(),
    "VOverExp": ViewHighlightImageProcessor(black_highlight_color=(70, 0, 70), white_highlight_color=(70, 70, 0)),
    "VAnnotat": ViewAnnotationsImageProcessor(base_image_selector='visualization'),
    "VCrosSec": ViewCrossSectionImageProcessor(
        centroid_pixel_locator, single_plot=False, y_range=(0, 255), base_image_selector='visualization'
    ),
    "VFalseC4": ViewHighlightImageProcessor(base_image_selector='visualization'),
    'Ve3d': View3dImageProcessor(max_resolution=(100, 100)),
}
_p = image_processors
processors_per_slide = [
    [
        (_p["Original"], "Original"),
        (_p["Rgb2Gray"], "Grayscale"),
        (_p["Centroid"], "Centroid & Principle Axis"),
        (_p["CropCent"], "Cropped to Centroid"),
    ],
    [
        (_p["VFalseC2"], "Cropped to Centroid"),
        (_p["VHotspot"], "Hotspot"),
        (_p["CropHots"], "Cropped to Hotspot"),
        (_p["VFalseC4"], "Test view over cross section"),
    ],
    [(_p["CropHots"], "Cropped Image"), _p["VAnnotat"], _p["VCrosSec"], _p["Ve3d"]],
]
image_processors["PowerPnt"] = PowerpointImageProcessor(
    results_dir, save_name_prefix, overwrite=True, processors_per_slide=processors_per_slide
)
image_processors_list = list(image_processors.values())

spot_analysis = SpotAnalysis(experiment_name, image_processors_list, save_dir=results_dir, save_overwrite=True)
spot_analysis.set_primary_images(input_images[:1])

for result in spot_analysis:
    pass

Without this fix, the resulting images for the "VCrosSec" and "VFalseC4" visualization image processors had the window dressings duplicated. After the fix the window dressings were no longer inherited by the later visualizations.

Submission checklist

• Target branch is develop, not main
• Existing tests are updated or new tests were added
• opencsp/test/test_DocStringsExist.py are verified to include this change or have been updated accordingly
• .rst file(s) under doc/ are verified to include this change or have been updated accordingly

Additional information

This needs additional work to add unit tests and update the documentation.

[jehsharp]

This looks great, thanks for fixing that bug, should it also include updates to the testing suite?