Merge pull request #1831 from kif/1825_worker_sigmaclip

Enable sigma_clip_legacy to be used within a worker

pyFAI/azimuthalIntegrator.py

@@ -30,7 +30,7 @@
 __contact__ = "[email protected]"
 __license__ = "MIT"
 __copyright__ = "European Synchrotron Radiation Facility, Grenoble, France"
-__date__ = "06/01/2023"
+__date__ = "22/02/2023"
 __status__ = "stable"
 __docformat__ = 'restructuredtext'
 
@@ -2899,21 +2899,21 @@ def medfilt1d(self, data, npt_rad=1024, npt_azim=512,
         return result
 
     def sigma_clip_legacy(self, data, npt_rad=1024, npt_azim=512,
-                           correctSolidAngle=True, polarization_factor=None,
-                           radial_range=None, azimuth_range=None,
-                           dark=None, flat=None,
-                           method="splitpixel", unit=units.Q,
-                           thres=3, max_iter=5, dummy=None, delta_dummy=None,
-                           mask=None, normalization_factor=1.0, metadata=None):
+                          correctSolidAngle=True, polarization_factor=None,
+                          radial_range=None, azimuth_range=None,
+                          dark=None, flat=None,
+                          method=("full", "histogram", "cython"), unit=units.Q,
+                          thres=3, max_iter=5, dummy=None, delta_dummy=None,
+                          mask=None, normalization_factor=1.0, metadata=None,
+                          safe=True, **kwargs):
         """Perform first a 2D integration and then an iterative sigma-clipping
         filter along each row. See the doc of scipy.stats.sigmaclip for the
         options `thres` and `max_iter`.
 
         :param data: input image as numpy array
-        :param npt_rad: number of radial points
+        :param npt_rad: number of radial points (alias: npt)
         :param npt_azim: number of azimuthal points
-        :param bool correctSolidAngle: correct for solid angle of each pixel
-                if True
+        :param bool correctSolidAngle: correct for solid angle of each pixel when set
         :param float polarization_factor: polarization factor between -1 (vertical)
                 and +1 (horizontal).
 
@@ -2928,17 +2928,22 @@ def sigma_clip_legacy(self, data, npt_rad=1024, npt_azim=512,
         :param ndarray flat: flat field image
         :param unit: unit to be used for integration
         :param method: pathway for integration and sort
-        :param thres: cut-off for n*sigma: discard any values with `|I-<I>|/sigma > thres`.
+        :param thres: cut-off for n*sigma: discard any values with `|I-<I>| > thres*σ`.
                 The threshold can be a 2-tuple with sigma_low and sigma_high.
         :param max_iter: maximum number of iterations
         :param mask: masked out pixels array
         :param float normalization_factor: Value of a normalization monitor
         :param metadata: any other metadata,
         :type metadata: JSON serializable dict
+        :param safe: unset to save some checks on sparse matrix shape/content.
         :return: Integrate1D-like result
 
         Nota: The initial 2D-integration requires pixel splitting
         """
+        #compatibility layer with sigma_clip_ng
+        if "npt" in kwargs:
+            npt_rad = kwargs["npt"]
+
         # We use NaN as dummies
         if dummy is None:
             dummy = numpy.NaN
@@ -2966,7 +2971,7 @@ def sigma_clip_legacy(self, data, npt_rad=1024, npt_azim=512,
                                  dummy=dummy, delta_dummy=delta_dummy,
                                  correctSolidAngle=correctSolidAngle,
                                  polarization_factor=polarization_factor,
-                                 normalization_factor=normalization_factor)
+                                 normalization_factor=normalization_factor,safe=safe)
         image = res2d.intensity
         if (method.impl_lower == "opencl"):
             if (method.algo_lower == "csr") and \
@@ -3079,7 +3084,7 @@ def sigma_clip_ng(self, data,
 
         Keep only pixels with intensty:
 
-            ``|I - <I>| < thres * std(I)``
+            ``|I - <I>| < thres * σ(I)``
 
         This enforces a symmetric, bell-shaped distibution (i.e. gaussian-like) and is very good at extracting
         background or amorphous isotropic scattering out of Bragg peaks.
@@ -3326,10 +3331,7 @@ def sigma_clip_ng(self, data,
         result._set_error_model(error_model)
         return result
 
-    @deprecated(reason="will be replaced by `sigma_clip_ng` in version 0.23.0. Please use either `_sigma_clip_legacy` for full compatibility or upgrade your code to accomodate the new API",
-                replacement="sigma_clip_ng", since_version="0.21.0", only_once=True, skip_backtrace_count=1, deprecated_since="0.22.0")
-    def sigma_clip(self, *args, **kwargs):
-        return self._sigma_clip_legacy(*args, **kwargs)
+    sigma_clip = sigma_clip_ng
 
     def separate(self, data, npt_rad=1024, npt_azim=512, unit="2th_deg", method="splitpixel",
                  percentile=50, mask=None, restore_mask=True):

pyFAI/test/test_worker.py

@@ -32,7 +32,7 @@
 __contact__ = "[email protected]"
 __license__ = "MIT"
 __copyright__ = "European Synchrotron Radiation Facility, Grenoble, France"
-__date__ = "26/11/2022"
+__date__ = "22/02/2023"
 
 import unittest
 import logging
@@ -307,8 +307,20 @@ def test_sigma_clip(self):
                         extra_options={"thres":2, "error_model":"azimuthal"},
                         integrator_name="sigma_clip_ng",
                         shapeOut=(1, 100))
+
         self.assertEqual(worker.error_model, "azimuthal", "error model is OK")
+        img = numpy.random.random(ai.detector.shape)
+        worker(img)
 
+    def test_sigma_clip_legacy(self):
+        "Non regression test for #1825"
+        ai = AzimuthalIntegrator.sload({"detector": "Pilatus100k", "wavelength":1e-10})
+        worker = Worker(azimuthalIntegrator=ai,
+                        extra_options={"thres":2},
+                        integrator_name="sigma_clip_legacy",
+                        shapeOut=(1, 100))
+        img = numpy.random.random(ai.detector.shape)
+        worker(img)
 
 class TestWorkerConfig(unittest.TestCase):