azav module

Azimuthal average related functions.

azav.get_mask_circle(img, radius, center=None)[source]

Return an image with a masked circle.

Parameters:

  • img (numpy array) – Input image.

  • radius (float) – Circle radius (in pixels).

  • center (tuple) – Center coordinates (x,y) (in pixels). If None, center=(nrows/2,ncols/2).

Returns:

mask – image mask

Return type:

numpy array

azav.get_mask_from_limits(img, mask_lim, invert=False)[source]

Get mask on the basis of a range defining tuple.

Parameters:

  • img (array_like) – Image to be analyzed.

  • mask_lim (tuple or list) – Pixels with intensity < mask_lim[0] or > mask_lim[1] will be True in the mask.

  • invert (bool) – If True, ‘mask’ is inverted. Default is False

Returns:

mask – Image mask.

Return type:

numpy array

azav.get_mask_zingers(img, threshold=20000.0, circle_radius=5.0, verbose=False)[source]

Returns a mask of the input image where pixels corresponding to zingers are masked with circles.

Zingers are clustered pixels with intensity >= threshold.

Parameters:

  • img (numpy array) – Input image.

  • threshold (float) – Pixel intensity threshold.

  • circle_radius (float) – Radius of circles used to mask zingers.

  • verbose (bool) – If True, information on zingers detection is printed out.

Returns:

mask – image mask

Return type:

numpy array

azav.get_poni(center, pixel, distance, tilt=0, tilt_plane_rotation=0)[source]

Get poni1, poni2, rot1, rot2 and rot3.

Parameters:

  • center (2D-tuple) – X-ray beam position on detector image (pixel).

  • pixel (2D-tuple) – Detector pixel size (m).

  • distance (float) – Sample to detector distance (m).

  • tilt (float) – Detector tilt (degrees).

  • tilt_plane_rotation (float) – Rotation of the tilt plane around the detector z-axis (degrees)

azav.get_zingers(img, threshold=20000.0, plot=True, clim='auto', verbose=True)[source]

Find clusters of pixels with intensity >= threshold.

The center-of-mass of the clustered is returned.

Parameters:

  • img (numpy array) – image to be analyzed

  • threshold (float) – pixel intensity threshold

  • plot (bool) – if True, a red circle is plot around each cluster

  • clim (str or tuple) – imshow clim, if “auto”, clim=(0,threshold)

  • verbose (bool) – if True, information on results are printed out

Returns:

cluster_xy – list of clusters center-of-mass

Return type:

list

azav.integrate1d(img, ai, npt=600, method='csr', unit='q_A^-1', normalization_factor=1.0, mask=None, dark='auto', flat=None, solid_angle=True, polarization_factor=0.99, variance=None, error_model=None, radial_range=None, azimuthal_range=None, dummy=None, delta_dummy=None, safe=True, metadata=None, dezinger_method='mask_zingers', dezinger=None, sensor_material=None, sensor_thickness=None, sensor_density=None, sample_material=None, sample_thickness=None, sample_density=None, verbose=False)[source]

Perform azimuthal integration of an image.

Parameters:

  • img (2d ndarray) – Detector Image.

  • ai (pyFAI.azimuthalIntegrator.AzimuthalIntegrator obj) – PyFAI azimuthal integrator.

  • npt (int, optional) – Number of radial bins (number of points in the output pattern). Default is 600.

  • method (str, optional) – PyFAI azimuthal integration method. Available choices are (pixel splitting | algorithm | implementation): “numpy” : no split | histogram | python “cython” : no split | histogram | cython “BBox” : bounding box | histogram | cython “splitpixel” : full | histogram | cython “lut” : bounding box | look-up table | cython “csr” : bounding box | compressed sparse row | cython “nosplit_csr” : no split | compressed sparse row | cython “full_csr” : full | compressed sparse row | cython “lut_ocl” : bounding box | look-up table | opencl “csr_ocl” : bounding box | compressed sparse row | opencl To specify the device: “csr_ocl_1,2”. No pixel splitting –> high noise on bins with few pixels. Bounding box splitting –> fast, but blurs a bit the signal. Pseudo pixel splitting –> compromise between speed and precision. Full pixel splitting –> slow, high precision. Default is ‘csr’.

  • unit (str, optional) – Output unit. Can be “q_nm^-1”, “q_A^-1”, “2th_deg”, “2th_rad”, “r_mm”. Default is “q_A^-1”.

  • normalization_factor (float or None, optional) – Value of a normalization monitor.

  • mask (ndarray or None, optional) – Array (same size as image) with 1 for masked pixels, and 0 for valid pixels. Default is None.

  • dark (ndarray or str or None, optional) – Dark noise image. If ‘auto’ (default) and ‘ai.detector’ is equal to ‘rayonix’, the constant offset 10 will be removed from ‘img’. Otherwise ‘dark’ is set to None.

  • flat (ndarray of None, optional) – Flat field image. Default is None.

  • solid_angle (bool, optional) – If True (default), correct for solid angle of each pixel.

  • polarization_factor (float or None, optional) – Polarization factor between -1 (vertical) and +1 (horizontal). 0 for circular polarization or random, If None, no correction is applied. Default is 0.99.

  • variance (ndarray or None, optional) – Array containing the variance of the data in ‘img’. If None (default), no error propagation is done.

  • error_model (str or None, optional) – If “poisson”: variance = I. If “azimuthal”: variance = (I-<I>)^2. Default is None.

  • radial_range ((float, float) or None, optional) – The lower and upper values of the radial unit. If None (default), range is simply (img.min(), img.max()) and values outside the range are ignored.

  • azimuthal_range ((float, float) or None, optional) – The lower and upper values of the azimuthal angle. If None (default), range is simply (img.min(), img.max()) and values outside the range are ignored.

  • dummy (ndarray or None, optional) – Value for dead/masked pixels.

  • delta_dummy (ndarray or None, optional) – Precision for dummy value.

  • safe (bool, optional) – If True, extra checks are performed to ensure LUT/CSR is valid. Default is False.

  • metadata (dict or None, optional) – JSON serializable object containing the metadata. Default is None.

  • dezinger_method (str, optional) – Can be ‘medfilt1d’ or ‘mask_zingers’. Default is ‘mask_zingers’. Ignored if ‘dezinger’ is None.

  • dezinger (float or tuple or str or None, optional) –

    • If None (default), no zinger is removed.

    • If ‘dezinger_method’ is ‘medfilt1d’: Percentile used for removing pixels corresponding to zingers (or Bragg peaks) from the amorphous part of the image. Percentile=90: 90% of pixels are retained. Avoid using percentile>95.

    • If tuple, specifies a region to average out.

    • If str, can be only ‘mask_zingers’ (currently it is the only implemented alternative to the pyFAI medifilt1() method).

    • If ‘dezinger_method’ is ‘mask_zingers’: Must be len=2 tuple. The first element is the pixel intensity threshold used to detect zingers (40e3 is a good default). The second element is the radius of the masking circle (5 pixels is a good default).

  • sensor_material (str or None, optional) – Material of which the detector sensitive area is made of. Can be expressed as chemical formula or name from materials list available in the python package ‘xraydb’. If ‘auto’, the chemical formula and thickness are deduced from ‘ai.detector’. Default is None.

  • sensor_thickness (float or None, optional) – Thickness of the detector sensitive area (m). Default is None. If both ‘sensor_material’ and ‘sensor_thickness’ are not None, the azimuthal averaged intensity is scaled (divided) by the (angular dependent) sensor absorption.

  • sensor_density (float or None, optional) – Density of the detector sensitive area (kg/m3). If None, the density is guessed automatically.

  • sample_material (str or None, optional) – Material of which the sample is made of. Can be expressed as chemical formula or name from materials list available in the python package ‘xraydb’. Default is None.

  • sample_thickness (float or None, optional) – Thickness of the sample (m). Default is None. If both ‘sample_material’ and ‘sample_thickness’ are not None, the azimuthal averaged intensity is scaled (divided) by the (angular dependent) sample transmission.

  • sensor_density – Density of the sample (g/cm3). If None, the density is guessed automatically.

Returns:

res – Result of PyFAI 1D integration. Attributes include “radial” (center position of radial bins), “intensity” (azimuthally integrated intensities) and “sigma” (errors on intensities).

Return type:

pyFAI.containers.Integrate1dResult obj

Notes

  • To get main results only, it is possible to use:

    q, i, e = integrate1d(…)

    or even:

    q, i = integrate1d(…)

  • Property ‘zingers’ add to ‘res’: could be None or bool.

azav.integrate1d_dataset(folder, ai=None, save_fname='id09_azav.h5', force=False, extension='h5', npt=600, method='csr', unit='q_A^-1', normalization_factor=1.0, mask=None, dark='auto', flat=None, solid_angle=True, polarization_factor=0.99, variance=None, error_model='poisson', radial_range=None, azimuthal_range=None, dummy=None, delta_dummy=None, safe=True, metadata=None, dezinger_method='mask_zingers', dezinger=None, sensor_material='auto', sensor_thickness=None, sensor_density=None, sample_material=None, sample_thickness=None, sample_density=None, store_all_masks=False, verbose=True)[source]

Perform azimuthal integration of a dataset and save results.

At difference with ‘integrate1d_multi’: - all images are assumed to be in the same folder and have the same format - results are returned as a single dict ‘res’ - res[‘q’] is a 1D numpy array - res[‘i’] is a 2D numpy array

Parameters:

  • folder (str) – Dataset path.

  • ai (pyFAI.azimuthalIntegrator.AzimuthalIntegrator obj or None, optional) – PyFAI azimuthal integrator.

  • mask (ndarray or None, optional) – Array (same size as image) with 1 for masked pixels, and 0 for valid pixels. Default is None.

  • save_fname (str or None, optional) – If not None, the result of the azimuthal integration is saved in the save folder where the dataset is, with filename ‘save_fname’. Default is ‘id09_azav.h5’.

  • force (bool, optional) – If False (default), the azimuthal integration is not performed and previously saved result (if existing) are returned. If True, or the file corresponding to ‘save_fname’ does not exist, the integration is performed.

  • dezinger (npt ...) – See ‘integrate1d’ docinfo.

  • store_all_mask (bool, optional) – If False (default), only the mask applied to the first image is stored. If True, a mask is stored for each image. This is useful when ‘dezinger’ is not None.

Returns:

ret – Results dictionary containing: q (1D ndarray), i (2D ndarray), e (2D ndarray) and info on azimuthal integration.

Return type:

dict

azav.integrate1d_multi(imgs, ai, npt=600, method='csr', unit='q_A^-1', normalization_factor=1.0, mask=None, dark='auto', flat=None, solid_angle=True, polarization_factor=0.99, variance=None, error_model=None, radial_range=None, azimuthal_range=None, dummy=None, delta_dummy=None, safe=True, metadata=None, dezinger_method='mask_zingers', dezinger=None, sensor_material=None, sensor_thickness=None, sensor_density=None, sample_material=None, sample_thickness=None, sample_density=None, return_info=False, store_all_masks=False, verbose=True)[source]

Perform the azimuthal integration of a series of images.

Parameters:

  • imgs (list or ndarray) – Detector images.

  • ai (pyFAI.azimuthalIntegrator.AzimuthalIntegrator obj) – PyFAI azimuthal integrator.

  • dezinger (npt ...) – See ‘integrate1d’ docinfo.

  • return_info (bool, optional) – If True, all azimuthal average parameters are stored and returned.

  • store_all_mask (bool, optional) – If False (default), only the mask applied to the first image is stored. If True, a mask is stored for each image. This is useful when ‘dezinger’ is not None.

Returns:

  • q (list) – List of q-vectors.

  • i (list) – List of intensity vectors.

  • e (list) – List of intensity error vectors. Returned only if ‘error_model’ is not None.

  • info (dict) – Info on azimuthal average parameters. Retured only if ‘return_info’ is True.