Source code for sunkit_image.utils.utils

This module contains a collection of functions of general utility.
import warnings

import numpy as np
from scipy.interpolate import RectBivariateSpline
from skimage import measure

import astropy.units as u
from import all_coordinates_from_map

__all__ = [

[docs]def equally_spaced_bins(inner_value=1, outer_value=2, nbins=100): """ Define a set of equally spaced bins between the specified inner and outer values. The inner value must be strictly less than the outer value. Parameters ---------- inner_value : `float` The inner value of the bins. outer_value : `float` The outer value of the bins. nbins : `int` Number of bins. Returns ------- `numpy.ndarray` An array of shape ``(2, nbins)`` containing the bin edges. """ if inner_value >= outer_value: raise ValueError("The inner value must be strictly less than the outer value.") if nbins <= 0: raise ValueError("The number of bins must be strictly greater than 0.") bin_edges = np.zeros((2, nbins)) bin_edges[0, :] = np.arange(0, nbins) bin_edges[1, :] = np.arange(1, nbins + 1) return inner_value + bin_edges * (outer_value - inner_value) / nbins
[docs]def bin_edge_summary(r, binfit): """ Return a summary of the bin edges. Parameters ---------- r : `numpy.ndarray` An array of bin edges of shape (2, nbins) where nbins is the number of bins. binfit : {'center' | 'left' | 'right'} How to summarize the bin edges. Returns ------- `numpy.ndarray` A one dimensional array of values that summarize the location of the bins. """ if r.ndim != 2: raise ValueError("The bin edges must be two-dimensional with shape (2, nbins).") if r.shape[0] != 2: raise ValueError("The bin edges must be two-dimensional with shape (2, nbins).") if binfit == "center": summary = 0.5 * (r[0, :] + r[1, :]) elif binfit == "left": summary = r[0, :] elif binfit == "right": summary = r[1, :] else: raise ValueError('Keyword "binfit" must have value "center", "left" or "right"') return summary
[docs]def find_pixel_radii(smap, scale=None): """ Find the distance of every pixel in a map from the center of the Sun. The answer is returned in units of solar radii. Parameters ---------- smap :`` A SunPy map. scale : {`None` | `astropy.units.Quantity`}, optional The radius of the Sun expressed in map units. For example, in typical helioprojective Cartesian maps the solar radius is expressed in units of arcseconds. If None then the map is queried for the scale. Returns ------- radii : `astropy.units.Quantity` An array the same shape as the input map. Each entry in the array gives the distance in solar radii of the pixel in the corresponding entry in the input map data. """ # Calculate the coordinates of every pixel. coords = all_coordinates_from_map(smap) # TODO: check that the returned coordinates are indeed helioprojective cartesian # Calculate the radii of every pixel in helioprojective Cartesian # coordinate distance units. radii = np.sqrt(coords.Tx**2 + coords.Ty**2) # Re-scale the output to solar radii if scale is None: return u.R_sun * (radii / smap.rsun_obs) else: return u.R_sun * (radii / scale)
[docs]def get_radial_intensity_summary(smap, radial_bin_edges, scale=None, summary=np.mean, **summary_kwargs): """ Get a summary statistic of the intensity in a map as a function of radius. Parameters ---------- smap : `` A SunPy map. radial_bin_edges : `astropy.units.Quantity` A two-dimensional array of bin edges of shape ``(2, nbins)`` where "nbins" is the number of bins. scale : {``None`` | `astropy.units.Quantity`}, optional A length scale against which radial distances are measured, expressed in the map spatial units. For example, in AIA helioprojective Cartesian maps a useful length scale is the solar radius and is expressed in units of arcseconds. summary : ``function``, optional A function that returns a summary statistic of the distribution of intensity, at a given radius, for example `numpy.std`. summary_kwargs :`dict`, optional Keywords applicable to the summary function. Returns ------- intensity summary : `numpy.ndarray` A summary statistic of the radial intensity in the bins defined by the bin edges. """ if scale is None: s = smap.rsun_obs else: s = scale # Get the radial distance of every pixel from the center of the Sun. map_r = find_pixel_radii(smap, scale=s).to(u.R_sun) # Number of radial bins nbins = radial_bin_edges.shape[1] # Upper and lower edges lower_edge = [map_r > radial_bin_edges[0, i].to(u.R_sun) for i in range(0, nbins)] upper_edge = [map_r < radial_bin_edges[1, i].to(u.R_sun) for i in range(0, nbins)] # Calculate the summary statistic in the radial bins. with warnings.catch_warnings(): # We want to ignore RuntimeWarning: Mean of empty slice warnings.simplefilter("ignore", category=RuntimeWarning) return np.asarray( [summary([lower_edge[i] * upper_edge[i]], **summary_kwargs) for i in range(0, nbins)] )
[docs]def reform2d(array, factor=1): """ Reform a 2d array by a given factor. Parameters ---------- array : `numpy.ndarray` 2d array to be reformed. factor : `int`, optional The array is going to be magnified by the factor. Default is 1. Returns ------- `numpy.ndarray` Reformed array. """ if not isinstance(factor, int): raise ValueError("Parameter 'factor' must be an integer!") if len(np.shape(array)) != 2: raise ValueError("Input array must be 2d!") if factor > 1: congridx = RectBivariateSpline(np.arange(0, array.shape[0]), np.arange(0, array.shape[1]), array, kx=1, ky=1) reformed_array = congridx(np.arange(0, array.shape[0], 1 / factor), np.arange(0, array.shape[1], 1 / factor)) return reformed_array else: return array
[docs]def points_in_poly(poly): """ Return polygon as grid of points inside polygon. Only works for polygons defined with points which are all integers. Parameters ---------- poly : `numpy.ndarray` N x 2 array which defines all points at the edge of a polygon. Returns ------- `numpy.ndarray` N x 2 array, all points within the polygon. """ if np.shape(poly)[1] != 2: raise ValueError("Polygon must be defined as a n x 2 array!") # Convert to integers poly = np.array(poly, dtype=int).tolist() xs, ys = zip(*poly) minx, maxx = min(xs), max(xs) miny, maxy = min(ys), max(ys) # New polygon with the staring point as [0, 0] newPoly = [(int(x - minx), int(y - miny)) for (x, y) in poly] mask = measure.grid_points_in_poly((round(maxx - minx) + 1, round(maxy - miny) + 1), newPoly) # All points in polygon points = [[x + minx, y + miny] for x, y in zip(*np.nonzero(mask))] # Add edge points if missing for p in poly: if p not in points: points.append(p) return points
[docs]def remove_duplicate(edge): """ Remove duplicated points in a the edge of a polygon. Parameters ---------- edge : `numpy.ndarray` N x 2 array which defines all points at the edge of a polygon. Returns ------- `numpy.ndarray` Same as edge, but with duplicated points removed. """ shape = np.shape(edge) if shape[1] != 2: raise ValueError("Polygon must be defined as a n x 2 array!") new_edge = [] for i in range(shape[0]): p = edge[i] if not isinstance(p, list): p = p.tolist() if p not in new_edge: new_edge.append(p) return new_edge
[docs]def calculate_gamma(pm, vel, pnorm, n): """ Calculate gamma values. Parameters ---------- pm : `numpy.ndarray` Vector from point "p" to point "m". vel : `numpy.ndarray` Velocity vector. pnorm : `numpy.ndarray` Mode of ``pm``. n : `int` Number of points. Returns ------- `numpy.ndarray` calculated gamma values for velocity vector vel References ---------- * Equation (8) in Laurent Graftieaux, Marc Michard and Nathalie Grosjean. Combining PIV, POD and vortex identification algorithms for the study of unsteady turbulent swirling flows. Meas. Sci. Technol. 12, 1422, 2001. ( * Equation (1) in Jiajia Liu, Chris Nelson, Robert Erdelyi. Automated Swirl Detection Algorithm (ASDA) and Its Application to Simulation and Observational Data. Astrophys. J., 872, 22, 2019. ( """ cross = np.cross(pm, vel) vel_norm = np.linalg.norm(vel, axis=2) sint = cross / (pnorm * vel_norm + 1e-10) return np.nansum(sint, axis=1) / n