orpheus-npcf 0.2.1__cp310-cp310-musllinux_1_2_x86_64.whl

orpheus/flat2dgrid.py

@@ -0,0 +1,68 @@
+import numpy as np
+from scipy.interpolate import RegularGridInterpolator
+
+__all__ = ["FlatPixelGrid_2D","FlatDataGrid_2D"]
+
+class FlatPixelGrid_2D(object):
+    
+    def __init__(self, start_1, start_2, npix_1, npix_2, dpix_1, dpix_2):
+        self.start_1 = start_1
+        self.start_2 = start_2
+        self.npix_1 = npix_1
+        self.npix_2 = npix_2
+        self.dpix_1 = dpix_1
+        self.dpix_2 = dpix_2
+        self.pix1_lbounds = self.start_1 + self.dpix_1*np.arange(self.npix_1)
+        self.pix2_lbounds = self.start_2 + self.dpix_2*np.arange(self.npix_2)
+        self.pix1_centers = self.pix1_lbounds + self.dpix_1/2.
+        self.pix2_centers = self.pix2_lbounds + self.dpix_2/2.
+        
+    def todatagrid(self, data):
+        return FlatDataGrid_2D(data, self.start_1, self.start_2, self.dpix_1, self.dpix_2)
+        
+    def _regrid(self, other, data):
+        assert(isinstance(other,FlatPixelGrid_2D)) 
+        assert(data.shape == (self.npix_2,self.npix_1))
+        data_int = RegularGridInterpolator((self.pix2_centers, self.pix1_centers), data,
+                                           method="linear", bounds_error=False, fill_value=0)
+        oc1, oc2 = np.meshgrid(other.pix2_centers,other.pix1_centers, indexing='ij')
+        centers_mapgrid = np.array([oc1,oc2]).reshape((2,oc1.shape[1]*oc1.shape[0])).transpose()
+        data_regridded = data_int(centers_mapgrid).reshape((oc1.shape[0],oc1.shape[1]))
+        return other.todatagrid(data_regridded)
+                
+        
+class FlatDataGrid_2D(FlatPixelGrid_2D):
+    
+    def __init__(self, data, start_1, start_2, dpix_1, dpix_2):
+        """
+        Convention: O (unmasked) --> 1 (fully masked)
+        """
+        super().__init__(start_1, start_2, data.shape[1], data.shape[0], dpix_1, dpix_2)
+        self.data = data
+        
+    def regrid(self, other_grid):
+        return super()._regrid(other_grid, self.data)
+    
+    def samplePoints(self, nbar, method='Poisson', rng=None):
+        """ Sample points within all pixels ==0 """
+        assert(method in ["Poisson"])
+        if rng is None:
+            rng = np.random.RandomState()
+        mask1_lo = self.start_1
+        mask1_hi = self.start_1 + self.npix_1*self.dpix_1
+        mask2_lo = self.start_2
+        mask2_hi = self.start_2 + self.npix_2*self.dpix_2
+        mask_ext_area = (mask1_hi-mask1_lo)*(mask2_hi-mask2_lo)
+        # Assumes .data is mask s.t. we only sample in unmasked region, indexed by zero.
+        if method=="Poisson":
+            ngal_rand = int(nbar*mask_ext_area)
+            rand_1 = mask1_lo + (mask1_hi-mask1_lo)*rng.rand(ngal_rand)
+            rand_2 = mask2_lo + (mask2_hi-mask2_lo)*rng.rand(ngal_rand)
+            rand_ipix = np.floor((rand_2-mask2_lo)/self.dpix_2).astype(int)*self.npix_1 + np.floor((rand_1-mask1_lo)/self.dpix_1).astype(int)
+            infoot = self.data.flatten()[rand_ipix]==0
+            rand_1 = rand_1[infoot]
+            rand_2 = rand_2[infoot]
+        # Assumes .data is density map
+        if method=="LinBias":
+            raise NotImplementedError
+        return rand_1, rand_2