From a25197edfdbeb82ae43d6c70aac1e0f69c2d9483 Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Fri, 6 May 2022 16:15:10 -0700
Subject: [PATCH 01/15] First cut implementation of projector in Fourier space.

---
 simSPI/linear_simulator/projector.py | 92 +++++++++++++++++++++++++---
 tests/test_projector.py              | 29 ++++++++-
 2 files changed, 110 insertions(+), 11 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index ad573f38..225cb535 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -1,6 +1,7 @@
 """Class to generate tomographic projection."""
 
 import torch
+import numpy as np
 from pytorch3d.transforms import Rotate
 
 
@@ -21,18 +22,89 @@ def __init__(self, config):
         super(Projector, self).__init__()
 
         self.config = config
-        self.vol = torch.rand([self.config.side_len] * 3, dtype=torch.float32)
-        lin_coords = torch.linspace(-1.0, 1.0, self.config.side_len)
-        [x, y, z] = torch.meshgrid(
-            [
-                lin_coords,
-            ]
-            * 3
-        )
-        coords = torch.stack([y, x, z], dim=-1)
-        self.register_buffer("vol_coords", coords.reshape(-1, 3))
+        self.space = config.space
+        
+        if self.space == "real":
+            self.vol = torch.rand([self.config.side_len] * 3, dtype=torch.float32)
+            lin_coords = torch.linspace(-1.0, 1.0, self.config.side_len)
+            [x, y, z] = torch.meshgrid(
+                [
+                    lin_coords,
+                ]
+                * 3
+            )
+            coords = torch.stack([y, x, z], dim=-1)
+            self.register_buffer("vol_coords", coords.reshape(-1, 3))
+        elif self.space == "fourier":
+            # Assume DC coefficient is at self.vol[n//2+1,n//2+1]
+            # this means that self.vol = fftshift(fft3(fftshift(real_vol)))
+            self.vol = torch.rand([self.config.side_len] * 3, dtype=torch.complex64)
+            freq_coords = torch.fft.fftfreq(self.config.side_len,dtype=torch.float32)
+            [x, y] = torch.meshgrid(
+                [
+                    freq_coords,
+                ]
+                * 2
+            )
+            coords = torch.stack([y, x], dim=-1)
+            self.register_buffer("vol_coords", coords.reshape(-1, 2))            
 
     def forward(self, rot_params, proj_axis=-1):
+        if self.space == "real":
+            return self._forward_real(rot_params,proj_axis)
+        elif self.space == "fourier":
+            return self._forward_fourier(rot_params)
+
+    def _forward_fourier(self, rot_params):
+        """Output the tomographic projection of the volume in Fourier space.
+
+        Take a slide through the Fourier space volume whose normal is
+        oriented according to rot_params.  The volume is assumed to be cube
+        represented in the fourier space.  The output image follows
+        (batch x channel x height x width) convention of pytorch.  Therefore,
+        a dummy channel dimension is added at the end to projection.
+
+        Parameters
+        ----------
+        rot_params: dict of type str to {tensor}
+            Dictionary containing parameters for rotation, with keys
+                rotmat: str map to tensor
+                    rotation matrix (batch_size x 3 x 3) to rotate the volume
+
+        Returns
+        -------
+        projection: tensor
+            Tensor containing tomographic projection
+            (batch_size x 1 x sidelen x sidelen)
+        """
+        
+        rotmat = rot_params["rotmat"]
+        batch_sz = rotmat.shape[0]
+        rot_vol_coords = 2*self.vol_coords.repeat((batch_sz,1,1)).bmm(rotmat[:,:2,:])
+        
+        projection = torch.empty((batch_sz,
+                                  self.config.side_len,
+                                  self.config.side_len), dtype=torch.complex64)
+        projection.real = torch.nn.functional.grid_sample(
+            self.vol.real.repeat((batch_sz, 1, 1, 1, 1)),
+            rot_vol_coords[:, None, None, :, :],
+            align_corners=True,
+        ).reshape(batch_sz,
+                  self.config.side_len,
+                  self.config.side_len)
+
+        projection.imag = torch.nn.functional.grid_sample(
+            self.vol.imag.repeat((batch_sz, 1, 1, 1, 1)),
+            rot_vol_coords[:, None, None, :, :],
+            align_corners=True,
+        ).reshape(batch_sz,
+                  self.config.side_len,
+                  self.config.side_len)
+
+        projection = projection[:, None, :, :]
+        return projection
+    
+    def _forward_real(self, rot_params, proj_axis=-1):
         """Output the tomographic projection of the volume.
 
         First rotate the volume and then sum it along an axis.
diff --git a/tests/test_projector.py b/tests/test_projector.py
index 9ed5d6e8..f1f6628b 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -1,6 +1,7 @@
 """Test function for projector module."""
 
 import numpy as np
+import torch
 
 from simSPI.linear_simulator.projector import Projector
 
@@ -40,6 +41,7 @@ def init_data(path):
         config_dict = saved_data["config_dict"]
     else:
         config_dict = {}
+    config_dict["space"] = "real"
     config = AttrDict(config_dict)
     return saved_data, config
 
@@ -63,11 +65,12 @@ def normalized_mse(a, b):
     return (a - b).pow(2).sum().sqrt() / a.pow(2).sum().sqrt()
 
 
-def test_projector():
+def test_projector_real():
     """Test accuracy of projector function."""
     path = "tests/data/projector_data.npy"
 
     saved_data, config = init_data(path)
+    config.space = "real"
     rot_params = saved_data["rot_params"]
     projector = Projector(config)
     projector.vol = saved_data["volume"]
@@ -75,3 +78,27 @@ def test_projector():
     out = projector(rot_params)
     error = normalized_mse(saved_data["projector_output"], out).item()
     assert (error < 0.01) == 1
+
+def test_projector_fourier():
+    """Test accuracy of projector function."""
+    path = "tests/data/projector_data.npy"
+
+    saved_data, config = init_data(path)
+    config.space = "fourier"
+    rot_params = saved_data["rot_params"]
+    projector = Projector(config)
+    print(saved_data["volume"])
+    projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(saved_data["volume"])))
+
+    out = projector(rot_params)
+    fft_proj_out = torch.fft.fftshift(torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))),dim=(2,3))
+    print(out.dtype)
+    print(fft_proj_out[0,0,0])
+    print(out[0,0,0])
+    print((fft_proj_out.real/out.real).median())
+    print(out.shape[0],np.sqrt(out.shape[0]),1.0/np.sqrt(out.shape[0]))
+    error_r = normalized_mse(fft_proj_out.real, out.real).item()
+    error_i = normalized_mse(fft_proj_out.imag, out.imag).item()
+    assert (error_r < 0.01) == 1
+    assert (error_i < 0.01) == 1
+

From ba7210572b6322c3d9ae8c4b08fb530470923a56 Mon Sep 17 00:00:00 2001
From: Roy <roy.lederman@yale.edu>
Date: Fri, 6 May 2022 20:00:13 -0400
Subject: [PATCH 02/15] scale grids after rotation

---
 simSPI/linear_simulator/projector.py | 48 +++++++++++++++++-----------
 1 file changed, 30 insertions(+), 18 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index 225cb535..b2c93eae 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -1,7 +1,7 @@
 """Class to generate tomographic projection."""
 
-import torch
 import numpy as np
+import torch
 from pytorch3d.transforms import Rotate
 
 
@@ -23,7 +23,7 @@ def __init__(self, config):
 
         self.config = config
         self.space = config.space
-        
+
         if self.space == "real":
             self.vol = torch.rand([self.config.side_len] * 3, dtype=torch.float32)
             lin_coords = torch.linspace(-1.0, 1.0, self.config.side_len)
@@ -34,12 +34,15 @@ def __init__(self, config):
                 * 3
             )
             coords = torch.stack([y, x, z], dim=-1)
+            # Rescale coordinates to [-1,1] to be compatible with torch.nn.functional.grid_sample
+            coords = 2 * coords
+
             self.register_buffer("vol_coords", coords.reshape(-1, 3))
         elif self.space == "fourier":
             # Assume DC coefficient is at self.vol[n//2+1,n//2+1]
             # this means that self.vol = fftshift(fft3(fftshift(real_vol)))
             self.vol = torch.rand([self.config.side_len] * 3, dtype=torch.complex64)
-            freq_coords = torch.fft.fftfreq(self.config.side_len,dtype=torch.float32)
+            freq_coords = torch.fft.fftfreq(self.config.side_len, dtype=torch.float32)
             [x, y] = torch.meshgrid(
                 [
                     freq_coords,
@@ -47,11 +50,11 @@ def __init__(self, config):
                 * 2
             )
             coords = torch.stack([y, x], dim=-1)
-            self.register_buffer("vol_coords", coords.reshape(-1, 2))            
+            self.register_buffer("vol_coords", coords.reshape(-1, 2))
 
     def forward(self, rot_params, proj_axis=-1):
         if self.space == "real":
-            return self._forward_real(rot_params,proj_axis)
+            return self._forward_real(rot_params, proj_axis)
         elif self.space == "fourier":
             return self._forward_fourier(rot_params)
 
@@ -77,33 +80,30 @@ def _forward_fourier(self, rot_params):
             Tensor containing tomographic projection
             (batch_size x 1 x sidelen x sidelen)
         """
-        
+
         rotmat = rot_params["rotmat"]
         batch_sz = rotmat.shape[0]
-        rot_vol_coords = 2*self.vol_coords.repeat((batch_sz,1,1)).bmm(rotmat[:,:2,:])
-        
-        projection = torch.empty((batch_sz,
-                                  self.config.side_len,
-                                  self.config.side_len), dtype=torch.complex64)
+        rot_vol_coords = self.vol_coords.repeat((batch_sz, 1, 1)).bmm(rotmat[:, :2, :])
+
+        projection = torch.empty(
+            (batch_sz, self.config.side_len, self.config.side_len),
+            dtype=torch.complex64,
+        )
         projection.real = torch.nn.functional.grid_sample(
             self.vol.real.repeat((batch_sz, 1, 1, 1, 1)),
             rot_vol_coords[:, None, None, :, :],
             align_corners=True,
-        ).reshape(batch_sz,
-                  self.config.side_len,
-                  self.config.side_len)
+        ).reshape(batch_sz, self.config.side_len, self.config.side_len)
 
         projection.imag = torch.nn.functional.grid_sample(
             self.vol.imag.repeat((batch_sz, 1, 1, 1, 1)),
             rot_vol_coords[:, None, None, :, :],
             align_corners=True,
-        ).reshape(batch_sz,
-                  self.config.side_len,
-                  self.config.side_len)
+        ).reshape(batch_sz, self.config.side_len, self.config.side_len)
 
         projection = projection[:, None, :, :]
         return projection
-    
+
     def _forward_real(self, rot_params, proj_axis=-1):
         """Output the tomographic projection of the volume.
 
@@ -132,6 +132,18 @@ def _forward_real(self, rot_params, proj_axis=-1):
         t = Rotate(rotmat, device=self.vol_coords.device)
         rot_vol_coords = t.transform_points(self.vol_coords.repeat(batch_sz, 1, 1))
 
+        # rescale the coordinates to be compatible with the edge alignment of torch.nn.functional.grid_sample
+        if 0 == self.config.side_len % 2:  # even case
+            rot_vol_coords = (
+                (rot_vol_coords + 1)
+                * (self.config.side_len + 1)
+                / (self.config.side_len)
+            ) - 1
+        else:  # odd case
+            rot_vol_coords = (
+                (rot_vol_coords) * (self.config.side_len + 1) / (self.config.side_len)
+            )
+
         rot_vol = torch.nn.functional.grid_sample(
             self.vol.repeat(batch_sz, 1, 1, 1, 1),
             rot_vol_coords[:, None, None, :, :],

From ada366e8f50d4cade424d62f5e6daaac82a73576 Mon Sep 17 00:00:00 2001
From: Roy <roy.lederman@yale.edu>
Date: Fri, 6 May 2022 20:33:21 -0400
Subject: [PATCH 03/15] scale grids factor fixed

---
 simSPI/linear_simulator/projector.py | 30 +++++++++++++++-------------
 tests/test_projector.py              | 11 ++++++++--
 2 files changed, 25 insertions(+), 16 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index b2c93eae..fe38af94 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -34,8 +34,6 @@ def __init__(self, config):
                 * 3
             )
             coords = torch.stack([y, x, z], dim=-1)
-            # Rescale coordinates to [-1,1] to be compatible with torch.nn.functional.grid_sample
-            coords = 2 * coords
 
             self.register_buffer("vol_coords", coords.reshape(-1, 3))
         elif self.space == "fourier":
@@ -50,6 +48,8 @@ def __init__(self, config):
                 * 2
             )
             coords = torch.stack([y, x], dim=-1)
+            # Rescale coordinates to [-1,1] to be compatible with torch.nn.functional.grid_sample
+            coords = 2 * coords
             self.register_buffer("vol_coords", coords.reshape(-1, 2))
 
     def forward(self, rot_params, proj_axis=-1):
@@ -83,8 +83,21 @@ def _forward_fourier(self, rot_params):
 
         rotmat = rot_params["rotmat"]
         batch_sz = rotmat.shape[0]
+        
         rot_vol_coords = self.vol_coords.repeat((batch_sz, 1, 1)).bmm(rotmat[:, :2, :])
-
+        
+        # rescale the coordinates to be compatible with the edge alignment of torch.nn.functional.grid_sample
+        if 0 == self.config.side_len % 2:  # even case
+            rot_vol_coords = (
+                (rot_vol_coords + 1)
+                * (self.config.side_len )
+                / (self.config.side_len-1)
+            ) - 1
+        else:  # odd case
+            rot_vol_coords = (
+                (rot_vol_coords) * (self.config.side_len ) / (self.config.side_len-1)
+            )
+            
         projection = torch.empty(
             (batch_sz, self.config.side_len, self.config.side_len),
             dtype=torch.complex64,
@@ -132,17 +145,6 @@ def _forward_real(self, rot_params, proj_axis=-1):
         t = Rotate(rotmat, device=self.vol_coords.device)
         rot_vol_coords = t.transform_points(self.vol_coords.repeat(batch_sz, 1, 1))
 
-        # rescale the coordinates to be compatible with the edge alignment of torch.nn.functional.grid_sample
-        if 0 == self.config.side_len % 2:  # even case
-            rot_vol_coords = (
-                (rot_vol_coords + 1)
-                * (self.config.side_len + 1)
-                / (self.config.side_len)
-            ) - 1
-        else:  # odd case
-            rot_vol_coords = (
-                (rot_vol_coords) * (self.config.side_len + 1) / (self.config.side_len)
-            )
 
         rot_vol = torch.nn.functional.grid_sample(
             self.vol.repeat(batch_sz, 1, 1, 1, 1),
diff --git a/tests/test_projector.py b/tests/test_projector.py
index f1f6628b..8655084b 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -86,10 +86,17 @@ def test_projector_fourier():
     saved_data, config = init_data(path)
     config.space = "fourier"
     rot_params = saved_data["rot_params"]
+    #rot_params["rotmat"].data[0]=torch.tensor([[1,0,0],[0,1,0],[0,0,1]])
+    #print(rot_params["rotmat"])
     projector = Projector(config)
-    print(saved_data["volume"])
+    #print(saved_data["volume"])
     projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(saved_data["volume"])))
-
+    
+    print(projector.vol.shape)
+    sz = projector.vol.shape[0]
+    print("vol_coords", projector.vol_coords)
+    print("Vol Center", projector.vol[sz//2,sz//2,sz//2])
+    
     out = projector(rot_params)
     fft_proj_out = torch.fft.fftshift(torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))),dim=(2,3))
     print(out.dtype)

From e854004bc37eec2010cd2a9e82a5411827630035 Mon Sep 17 00:00:00 2001
From: Roy <roy.lederman@yale.edu>
Date: Fri, 6 May 2022 22:02:42 -0400
Subject: [PATCH 04/15] redo tests

---
 simSPI/linear_simulator/projector.py |  5 ++
 tests/test_projector.py              | 86 +++++++++++++++++++++++++---
 2 files changed, 84 insertions(+), 7 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index fe38af94..6628e027 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -84,7 +84,12 @@ def _forward_fourier(self, rot_params):
         rotmat = rot_params["rotmat"]
         batch_sz = rotmat.shape[0]
         
+        print(rotmat[0])
+        rotmat = torch.transpose(rotmat,-1,-2)
+        print(rotmat[0])
         rot_vol_coords = self.vol_coords.repeat((batch_sz, 1, 1)).bmm(rotmat[:, :2, :])
+        print(rot_vol_coords[0])
+        print(rot_vol_coords[1])
         
         # rescale the coordinates to be compatible with the edge alignment of torch.nn.functional.grid_sample
         if 0 == self.config.side_len % 2:  # even case
diff --git a/tests/test_projector.py b/tests/test_projector.py
index 8655084b..2ca96ddb 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -79,10 +79,73 @@ def test_projector_real():
     error = normalized_mse(saved_data["projector_output"], out).item()
     assert (error < 0.01) == 1
 
+
+def test_projector_fourier_axis_aligned():
+    """Test accuracy of projector function - simplified."""
+    path = "tests/data/projector_data.npy"
+
+    saved_data, config = init_data(path)
+    config.space = "fourier"
+    rot_params = saved_data["rot_params"]
+    rot_params["rotmat"].data[0]=torch.tensor([[1,0,0],[0,1,0],[0,0,1]])
+    rot_params["rotmat"].data[1]=torch.tensor([[0,1,0],[-1,0,0],[0,0,1]])
+    rot_params["rotmat"].data[2]=torch.tensor([[0,0,1],[0,1,0],[-1,0,0]])
+    #print(rot_params["rotmat"])
+    config["side_len"]=4
+    projector = Projector(config)
+    #print(saved_data["volume"])
+    vol = torch.fft.fftshift(saved_data["volume"],dim=[-3,-2,-1])
+    #vol_shift = vol
+    #vol_shift = vol
+    vol = torch.zeros((4,4,4),dtype=torch.float32)
+    vol[0,1,0]=1
+    vol_shift = vol
+    
+    #projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(vol_shift)))
+    projector.vol = torch.fft.ifftshift(torch.fft.fftn(vol_shift),dim=[-3,-2,-1])
+    
+    #print(projector.vol.shape)
+    sz = projector.vol.shape[0]
+    #print("vol_coords", projector.vol_coords)
+    #print("Vol Center", projector.vol[sz//2,sz//2,sz//2])
+    
+    out = projector(rot_params)
+    out_r = torch.fft.ifft2(out,dim=(2,3))
+    #fft_proj_out = torch.fft.fftshift(torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))),dim=(2,3))
+    #fft_proj_out = (torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))))
+    print("FOURIER2",out.dtype)
+    
+    print(vol_shift.shape)
+    
+    sum1 = vol_shift.sum(axis=1)
+    sum2 = vol_shift.sum(axis=2)
+    sum0 = vol_shift.sum(axis=0)
+    
+    print("sum0",sum0)
+    print("sum1",sum1)
+    print("sum2",sum2)
+    print(out_r[0])
+    print(out_r[1])
+    print(out_r[2])
+    
+    #print( sum1 )
+    #print( (sum0 - out_r[0]).numpy() )
+    #print( (sum1 - out_r[0]).numpy() 
+    print( (sum0 - out_r[0]).abs().pow(2).sum().numpy()  )
+    print( (sum1 - out_r[1]).abs().pow(2).sum().numpy() )
+    print( (sum2 - out_r[2]).abs().pow(2).sum().numpy() )
+    assert( (sum0 - out_r[0]).abs().pow(2).sum().numpy() < 1e-12 )
+    assert( (sum1 - out_r[1]).abs().pow(2).sum().numpy() < 1e-12 )
+    assert( (sum2 - out_r[2]).abs().pow(2).sum().numpy() < 1e-12 )
+    
+    assert( False )
+    
+
 def test_projector_fourier():
     """Test accuracy of projector function."""
     path = "tests/data/projector_data.npy"
-
+    
+    return 
     saved_data, config = init_data(path)
     config.space = "fourier"
     rot_params = saved_data["rot_params"]
@@ -90,22 +153,31 @@ def test_projector_fourier():
     #print(rot_params["rotmat"])
     projector = Projector(config)
     #print(saved_data["volume"])
-    projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(saved_data["volume"])))
+    projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(saved_data["volume"],dim=[-3,-2,-1])),dim=[-3,-2,-1])
     
-    print(projector.vol.shape)
+    #print(projector.vol.shape)
     sz = projector.vol.shape[0]
-    print("vol_coords", projector.vol_coords)
-    print("Vol Center", projector.vol[sz//2,sz//2,sz//2])
+    #print("vol_coords", projector.vol_coords)
+    #print("Vol Center", projector.vol[sz//2,sz//2,sz//2])
     
     out = projector(rot_params)
-    fft_proj_out = torch.fft.fftshift(torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))),dim=(2,3))
+    #fft_proj_out = torch.fft.fftshift(torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))),dim=(2,3))
+    fft_proj_out = (torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))))
     print(out.dtype)
     print(fft_proj_out[0,0,0])
     print(out[0,0,0])
-    print((fft_proj_out.real/out.real).median())
+    print("ratio", sz, (fft_proj_out.real/out.real).median())
+    print("ratio", sz, 1/(fft_proj_out.real[0,0,0,0]/out.real[0,0,0,0]))
+    print("ratio", sz, 1/(fft_proj_out.real[:,0,0,0]/out.real[:,0,0,0]))
     print(out.shape[0],np.sqrt(out.shape[0]),1.0/np.sqrt(out.shape[0]))
     error_r = normalized_mse(fft_proj_out.real, out.real).item()
     error_i = normalized_mse(fft_proj_out.imag, out.imag).item()
     assert (error_r < 0.01) == 1
     assert (error_i < 0.01) == 1
 
+
+
+
+
+    
+

From 062d0dd7dfc49d55239d81eb2c09781dd4ae2bf6 Mon Sep 17 00:00:00 2001
From: Roy <roy.lederman@yale.edu>
Date: Tue, 10 May 2022 15:24:35 -0400
Subject: [PATCH 05/15] test fft scale

---
 simSPI/linear_simulator/projector.py | 27 +++++++--
 tests/test_projector.py              | 87 ++++------------------------
 2 files changed, 34 insertions(+), 80 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index 6628e027..fbe4d101 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -77,19 +77,30 @@ def _forward_fourier(self, rot_params):
         Returns
         -------
         projection: tensor
-            Tensor containing tomographic projection
+            Tensor containing tomographic projection in the Fourier domain
             (batch_size x 1 x sidelen x sidelen)
+        
+        Comments
+        --------
+        Note that the Fourier volumes are arbitrary channel x height x width complex valued tensors, 
+        they are not assumed to be Fourier transforms of a real valued 3D functions.
+        
+        Note that the tomographic projection is interpolated on a rotated 2D grid. 
+        The rotated 2D grid extends outside the boundaries of the 3D grid. 
+        The values outside the boundaries are not defined in a useful way. 
+        Therefore, in most applications, it make sense to apply a radial filter to the sample.
+
         """
 
         rotmat = rot_params["rotmat"]
         batch_sz = rotmat.shape[0]
         
-        print(rotmat[0])
+        #print(rotmat[0])
         rotmat = torch.transpose(rotmat,-1,-2)
-        print(rotmat[0])
+        #print(rotmat[0])
         rot_vol_coords = self.vol_coords.repeat((batch_sz, 1, 1)).bmm(rotmat[:, :2, :])
-        print(rot_vol_coords[0])
-        print(rot_vol_coords[1])
+        #print(rot_vol_coords[0])
+        #print(rot_vol_coords[1])
         
         # rescale the coordinates to be compatible with the edge alignment of torch.nn.functional.grid_sample
         if 0 == self.config.side_len % 2:  # even case
@@ -107,16 +118,22 @@ def _forward_fourier(self, rot_params):
             (batch_sz, self.config.side_len, self.config.side_len),
             dtype=torch.complex64,
         )
+        # interpolation is decomposed to real and imaginary parts due to torch grid_sample type rules. Requires data and coordinates of same type.
+        # padding_mode="reflection" is required due to possible pathologies right on the border.
+        # however, padding_mode="zeros" is what users might expect in most cases other than these axis aligned cases.
+        padding_mode="zeros"
         projection.real = torch.nn.functional.grid_sample(
             self.vol.real.repeat((batch_sz, 1, 1, 1, 1)),
             rot_vol_coords[:, None, None, :, :],
             align_corners=True,
+            padding_mode=padding_mode,
         ).reshape(batch_sz, self.config.side_len, self.config.side_len)
 
         projection.imag = torch.nn.functional.grid_sample(
             self.vol.imag.repeat((batch_sz, 1, 1, 1, 1)),
             rot_vol_coords[:, None, None, :, :],
             align_corners=True,
+            padding_mode=padding_mode,
         ).reshape(batch_sz, self.config.side_len, self.config.side_len)
 
         projection = projection[:, None, :, :]
diff --git a/tests/test_projector.py b/tests/test_projector.py
index 2ca96ddb..3c70a713 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -80,100 +80,37 @@ def test_projector_real():
     assert (error < 0.01) == 1
 
 
-def test_projector_fourier_axis_aligned():
-    """Test accuracy of projector function - simplified."""
-    path = "tests/data/projector_data.npy"
-
-    saved_data, config = init_data(path)
-    config.space = "fourier"
-    rot_params = saved_data["rot_params"]
-    rot_params["rotmat"].data[0]=torch.tensor([[1,0,0],[0,1,0],[0,0,1]])
-    rot_params["rotmat"].data[1]=torch.tensor([[0,1,0],[-1,0,0],[0,0,1]])
-    rot_params["rotmat"].data[2]=torch.tensor([[0,0,1],[0,1,0],[-1,0,0]])
-    #print(rot_params["rotmat"])
-    config["side_len"]=4
-    projector = Projector(config)
-    #print(saved_data["volume"])
-    vol = torch.fft.fftshift(saved_data["volume"],dim=[-3,-2,-1])
-    #vol_shift = vol
-    #vol_shift = vol
-    vol = torch.zeros((4,4,4),dtype=torch.float32)
-    vol[0,1,0]=1
-    vol_shift = vol
-    
-    #projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(vol_shift)))
-    projector.vol = torch.fft.ifftshift(torch.fft.fftn(vol_shift),dim=[-3,-2,-1])
-    
-    #print(projector.vol.shape)
-    sz = projector.vol.shape[0]
-    #print("vol_coords", projector.vol_coords)
-    #print("Vol Center", projector.vol[sz//2,sz//2,sz//2])
-    
-    out = projector(rot_params)
-    out_r = torch.fft.ifft2(out,dim=(2,3))
-    #fft_proj_out = torch.fft.fftshift(torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))),dim=(2,3))
-    #fft_proj_out = (torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))))
-    print("FOURIER2",out.dtype)
-    
-    print(vol_shift.shape)
-    
-    sum1 = vol_shift.sum(axis=1)
-    sum2 = vol_shift.sum(axis=2)
-    sum0 = vol_shift.sum(axis=0)
-    
-    print("sum0",sum0)
-    print("sum1",sum1)
-    print("sum2",sum2)
-    print(out_r[0])
-    print(out_r[1])
-    print(out_r[2])
-    
-    #print( sum1 )
-    #print( (sum0 - out_r[0]).numpy() )
-    #print( (sum1 - out_r[0]).numpy() 
-    print( (sum0 - out_r[0]).abs().pow(2).sum().numpy()  )
-    print( (sum1 - out_r[1]).abs().pow(2).sum().numpy() )
-    print( (sum2 - out_r[2]).abs().pow(2).sum().numpy() )
-    assert( (sum0 - out_r[0]).abs().pow(2).sum().numpy() < 1e-12 )
-    assert( (sum1 - out_r[1]).abs().pow(2).sum().numpy() < 1e-12 )
-    assert( (sum2 - out_r[2]).abs().pow(2).sum().numpy() < 1e-12 )
-    
-    assert( False )
     
 
 def test_projector_fourier():
-    """Test accuracy of projector function."""
+    """Test accuracy of projector function.
+    Note: corrent test only checks that the scaling is compatible.
+    """
+    
     path = "tests/data/projector_data.npy"
     
-    return 
     saved_data, config = init_data(path)
     config.space = "fourier"
     rot_params = saved_data["rot_params"]
-    #rot_params["rotmat"].data[0]=torch.tensor([[1,0,0],[0,1,0],[0,0,1]])
-    #print(rot_params["rotmat"])
     projector = Projector(config)
-    #print(saved_data["volume"])
     projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(saved_data["volume"],dim=[-3,-2,-1])),dim=[-3,-2,-1])
     
-    #print(projector.vol.shape)
     sz = projector.vol.shape[0]
-    #print("vol_coords", projector.vol_coords)
-    #print("Vol Center", projector.vol[sz//2,sz//2,sz//2])
     
     out = projector(rot_params)
-    #fft_proj_out = torch.fft.fftshift(torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))),dim=(2,3))
     fft_proj_out = (torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))))
+    #for j in range(fft_proj_out.shape[0]):
+    #    print("image relative error ", j, (fft_proj_out[j]-out[j]).norm()/fft_proj_out[j].norm())
     print(out.dtype)
-    print(fft_proj_out[0,0,0])
-    print(out[0,0,0])
     print("ratio", sz, (fft_proj_out.real/out.real).median())
     print("ratio", sz, 1/(fft_proj_out.real[0,0,0,0]/out.real[0,0,0,0]))
     print("ratio", sz, 1/(fft_proj_out.real[:,0,0,0]/out.real[:,0,0,0]))
-    print(out.shape[0],np.sqrt(out.shape[0]),1.0/np.sqrt(out.shape[0]))
-    error_r = normalized_mse(fft_proj_out.real, out.real).item()
-    error_i = normalized_mse(fft_proj_out.imag, out.imag).item()
-    assert (error_r < 0.01) == 1
-    assert (error_i < 0.01) == 1
+    assert( 0.01 > (fft_proj_out.real[0,0,0,0]/out.real[0,0,0,0]-1).abs() )
+    #print(out.shape[0],np.sqrt(out.shape[0]),1.0/np.sqrt(out.shape[0]))
+    #error_r = normalized_mse(fft_proj_out.real, out.real).item()
+    #error_i = normalized_mse(fft_proj_out.imag, out.imag).item()
+    #assert (error_r < 0.01) == 1
+    #assert (error_i < 0.01) == 1
 
 
 

From fde4f09c4a2ea1d4bbc3f0800c7ecdbbc9481555 Mon Sep 17 00:00:00 2001
From: Roy <lederman@users.noreply.github.com>
Date: Tue, 10 May 2022 17:22:30 -0400
Subject: [PATCH 06/15] style corrections

---
 simSPI/linear_simulator/projector.py | 49 +++++++++++++++-------------
 tests/test_projector.py              | 46 ++++++++++++--------------
 2 files changed, 48 insertions(+), 47 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index fbe4d101..60740f1f 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -1,6 +1,5 @@
 """Class to generate tomographic projection."""
 
-import numpy as np
 import torch
 from pytorch3d.transforms import Rotate
 
@@ -48,11 +47,18 @@ def __init__(self, config):
                 * 2
             )
             coords = torch.stack([y, x], dim=-1)
-            # Rescale coordinates to [-1,1] to be compatible with torch.nn.functional.grid_sample
+            # Rescale coordinates to [-1,1] to be compatible with
+            # torch.nn.functional.grid_sample
             coords = 2 * coords
             self.register_buffer("vol_coords", coords.reshape(-1, 2))
 
     def forward(self, rot_params, proj_axis=-1):
+        """Forward method for projection.
+
+        Parameters
+        ----------
+        rot_params : tensor of rotation matrices
+        """
         if self.space == "real":
             return self._forward_real(rot_params, proj_axis)
         elif self.space == "fourier":
@@ -79,41 +85,41 @@ def _forward_fourier(self, rot_params):
         projection: tensor
             Tensor containing tomographic projection in the Fourier domain
             (batch_size x 1 x sidelen x sidelen)
-        
+
         Comments
         --------
-        Note that the Fourier volumes are arbitrary channel x height x width complex valued tensors, 
+        Note that the Fourier volumes are arbitrary 
+        channel x height x width complex valued tensors,
         they are not assumed to be Fourier transforms of a real valued 3D functions.
-        
-        Note that the tomographic projection is interpolated on a rotated 2D grid. 
-        The rotated 2D grid extends outside the boundaries of the 3D grid. 
-        The values outside the boundaries are not defined in a useful way. 
+
+        Note that the tomographic projection is interpolated on a rotated 2D grid.
+        The rotated 2D grid extends outside the boundaries of the 3D grid.
+        The values outside the boundaries are not defined in a useful way.
         Therefore, in most applications, it make sense to apply a radial filter to the sample.
 
         """
-
         rotmat = rot_params["rotmat"]
         batch_sz = rotmat.shape[0]
-        
-        #print(rotmat[0])
-        rotmat = torch.transpose(rotmat,-1,-2)
-        #print(rotmat[0])
+
+        # print(rotmat[0])
+        rotmat = torch.transpose(rotmat, -1, -2)
+        # print(rotmat[0])
         rot_vol_coords = self.vol_coords.repeat((batch_sz, 1, 1)).bmm(rotmat[:, :2, :])
-        #print(rot_vol_coords[0])
-        #print(rot_vol_coords[1])
-        
+        # print(rot_vol_coords[0])
+        # print(rot_vol_coords[1])
+
         # rescale the coordinates to be compatible with the edge alignment of torch.nn.functional.grid_sample
         if 0 == self.config.side_len % 2:  # even case
             rot_vol_coords = (
                 (rot_vol_coords + 1)
-                * (self.config.side_len )
-                / (self.config.side_len-1)
+                * (self.config.side_len)
+                / (self.config.side_len - 1)
             ) - 1
         else:  # odd case
             rot_vol_coords = (
-                (rot_vol_coords) * (self.config.side_len ) / (self.config.side_len-1)
+                (rot_vol_coords) * (self.config.side_len) / (self.config.side_len - 1)
             )
-            
+
         projection = torch.empty(
             (batch_sz, self.config.side_len, self.config.side_len),
             dtype=torch.complex64,
@@ -121,7 +127,7 @@ def _forward_fourier(self, rot_params):
         # interpolation is decomposed to real and imaginary parts due to torch grid_sample type rules. Requires data and coordinates of same type.
         # padding_mode="reflection" is required due to possible pathologies right on the border.
         # however, padding_mode="zeros" is what users might expect in most cases other than these axis aligned cases.
-        padding_mode="zeros"
+        padding_mode = "zeros"
         projection.real = torch.nn.functional.grid_sample(
             self.vol.real.repeat((batch_sz, 1, 1, 1, 1)),
             rot_vol_coords[:, None, None, :, :],
@@ -167,7 +173,6 @@ def _forward_real(self, rot_params, proj_axis=-1):
         t = Rotate(rotmat, device=self.vol_coords.device)
         rot_vol_coords = t.transform_points(self.vol_coords.repeat(batch_sz, 1, 1))
 
-
         rot_vol = torch.nn.functional.grid_sample(
             self.vol.repeat(batch_sz, 1, 1, 1, 1),
             rot_vol_coords[:, None, None, :, :],
diff --git a/tests/test_projector.py b/tests/test_projector.py
index 3c70a713..e5b095ec 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -80,41 +80,37 @@ def test_projector_real():
     assert (error < 0.01) == 1
 
 
-    
-
 def test_projector_fourier():
     """Test accuracy of projector function.
     Note: corrent test only checks that the scaling is compatible.
     """
-    
+
     path = "tests/data/projector_data.npy"
-    
+
     saved_data, config = init_data(path)
     config.space = "fourier"
     rot_params = saved_data["rot_params"]
     projector = Projector(config)
-    projector.vol = torch.fft.fftshift(torch.fft.fftn(torch.fft.fftshift(saved_data["volume"],dim=[-3,-2,-1])),dim=[-3,-2,-1])
-    
+    projector.vol = torch.fft.fftshift(
+        torch.fft.fftn(torch.fft.fftshift(saved_data["volume"], dim=[-3, -2, -1])),
+        dim=[-3, -2, -1],
+    )
+
     sz = projector.vol.shape[0]
-    
+
     out = projector(rot_params)
-    fft_proj_out = (torch.fft.fft2(torch.fft.fftshift(saved_data["projector_output"],dim=(2,3))))
-    #for j in range(fft_proj_out.shape[0]):
+    fft_proj_out = torch.fft.fft2(
+        torch.fft.fftshift(saved_data["projector_output"], dim=(2, 3))
+    )
+    # for j in range(fft_proj_out.shape[0]):
     #    print("image relative error ", j, (fft_proj_out[j]-out[j]).norm()/fft_proj_out[j].norm())
     print(out.dtype)
-    print("ratio", sz, (fft_proj_out.real/out.real).median())
-    print("ratio", sz, 1/(fft_proj_out.real[0,0,0,0]/out.real[0,0,0,0]))
-    print("ratio", sz, 1/(fft_proj_out.real[:,0,0,0]/out.real[:,0,0,0]))
-    assert( 0.01 > (fft_proj_out.real[0,0,0,0]/out.real[0,0,0,0]-1).abs() )
-    #print(out.shape[0],np.sqrt(out.shape[0]),1.0/np.sqrt(out.shape[0]))
-    #error_r = normalized_mse(fft_proj_out.real, out.real).item()
-    #error_i = normalized_mse(fft_proj_out.imag, out.imag).item()
-    #assert (error_r < 0.01) == 1
-    #assert (error_i < 0.01) == 1
-
-
-
-
-
-    
-
+    print("ratio", sz, (fft_proj_out.real / out.real).median())
+    print("ratio", sz, 1 / (fft_proj_out.real[0, 0, 0, 0] / out.real[0, 0, 0, 0]))
+    print("ratio", sz, 1 / (fft_proj_out.real[:, 0, 0, 0] / out.real[:, 0, 0, 0]))
+    assert 0.01 > (fft_proj_out.real[0, 0, 0, 0] / out.real[0, 0, 0, 0] - 1).abs()
+    # print(out.shape[0],np.sqrt(out.shape[0]),1.0/np.sqrt(out.shape[0]))
+    # error_r = normalized_mse(fft_proj_out.real, out.real).item()
+    # error_i = normalized_mse(fft_proj_out.imag, out.imag).item()
+    # assert (error_r < 0.01) == 1
+    # assert (error_i < 0.01) == 1

From de7c74655b484eeebf786646646a8ca4b0071b7e Mon Sep 17 00:00:00 2001
From: Roy <lederman@users.noreply.github.com>
Date: Tue, 10 May 2022 17:32:33 -0400
Subject: [PATCH 07/15] style corrections

---
 simSPI/linear_simulator/projector.py | 17 +++++++++++------
 tests/test_projector.py              |  5 ++---
 2 files changed, 13 insertions(+), 9 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index 60740f1f..b83aa252 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -88,14 +88,15 @@ def _forward_fourier(self, rot_params):
 
         Comments
         --------
-        Note that the Fourier volumes are arbitrary 
+        Note that the Fourier volumes are arbitrary
         channel x height x width complex valued tensors,
         they are not assumed to be Fourier transforms of a real valued 3D functions.
 
         Note that the tomographic projection is interpolated on a rotated 2D grid.
         The rotated 2D grid extends outside the boundaries of the 3D grid.
         The values outside the boundaries are not defined in a useful way.
-        Therefore, in most applications, it make sense to apply a radial filter to the sample.
+        Therefore, in most applications, it make sense to apply a radial filter
+        to the sample.
 
         """
         rotmat = rot_params["rotmat"]
@@ -108,7 +109,8 @@ def _forward_fourier(self, rot_params):
         # print(rot_vol_coords[0])
         # print(rot_vol_coords[1])
 
-        # rescale the coordinates to be compatible with the edge alignment of torch.nn.functional.grid_sample
+        # rescale the coordinates to be compatible with the edge alignment of
+        # torch.nn.functional.grid_sample
         if 0 == self.config.side_len % 2:  # even case
             rot_vol_coords = (
                 (rot_vol_coords + 1)
@@ -124,9 +126,12 @@ def _forward_fourier(self, rot_params):
             (batch_sz, self.config.side_len, self.config.side_len),
             dtype=torch.complex64,
         )
-        # interpolation is decomposed to real and imaginary parts due to torch grid_sample type rules. Requires data and coordinates of same type.
-        # padding_mode="reflection" is required due to possible pathologies right on the border.
-        # however, padding_mode="zeros" is what users might expect in most cases other than these axis aligned cases.
+        # interpolation is decomposed to real and imaginary parts due to torch
+        # grid_sample type rules. Requires data and coordinates of same type.
+        # padding_mode="reflection" is required due to possible pathologies
+        # right on the border.
+        # however, padding_mode="zeros" is what users might expect in most
+        # cases other than these axis aligned cases.
         padding_mode = "zeros"
         projection.real = torch.nn.functional.grid_sample(
             self.vol.real.repeat((batch_sz, 1, 1, 1, 1)),
diff --git a/tests/test_projector.py b/tests/test_projector.py
index e5b095ec..2bbb7f8f 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -82,9 +82,9 @@ def test_projector_real():
 
 def test_projector_fourier():
     """Test accuracy of projector function.
+
     Note: corrent test only checks that the scaling is compatible.
     """
-
     path = "tests/data/projector_data.npy"
 
     saved_data, config = init_data(path)
@@ -102,8 +102,7 @@ def test_projector_fourier():
     fft_proj_out = torch.fft.fft2(
         torch.fft.fftshift(saved_data["projector_output"], dim=(2, 3))
     )
-    # for j in range(fft_proj_out.shape[0]):
-    #    print("image relative error ", j, (fft_proj_out[j]-out[j]).norm()/fft_proj_out[j].norm())
+
     print(out.dtype)
     print("ratio", sz, (fft_proj_out.real / out.real).median())
     print("ratio", sz, 1 / (fft_proj_out.real[0, 0, 0, 0] / out.real[0, 0, 0, 0]))

From 92644f4e2ddb2a4f2dd1c2e4cd2cb33c35238293 Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 10:48:10 -0400
Subject: [PATCH 08/15] Update linear_simulator config

---
 tests/data/linear_simulator_data.npy      | Bin 300145 -> 300168 bytes
 tests/data/linear_simulator_data_cube.npy | Bin 300164 -> 300187 bytes
 2 files changed, 0 insertions(+), 0 deletions(-)

diff --git a/tests/data/linear_simulator_data.npy b/tests/data/linear_simulator_data.npy
index db745d1805012a25adc57e06f52c27cfbad3f178..63eaac2b0f7fc8af448a560e5607ecf2f000cf97 100644
GIT binary patch
delta 1207
zcmY+D%TE(g6vn61;=3T|!VL)<NHvkv;Z-!+2cqDBkE`hTtV3yO721C79Rvlf;wUH}
zSha%ho$ib*x-+;Tk;q=c|6t-jpx)_)Ix~;td^eNxJLi07et)bO`&9AyfVfF&ZfvTn
zcYCreVs%l>C&e!|uIeODc8{Z`wD27hu&K2|W?pqPERJTE^4TTZ&;9&Msos<OyTega
zRQV4jBK03sN(4bjC)H3GUd6%UR9NkUPgAzAI7B1Cq?%Iu)9{ZV;8HfT!=7*?5z^8@
z4K2BuHgS<>+}G&IeJizBhYF=IWt3v%Gp^+y6W&N<gHI3^ymPZFUL4lZN^nFLT(U#9
z88^jI7mtj%Z9KYbH$gkaF#_3w9YBW(>-KOgFTzk0?P6u1Q^#?F6PElZP5Dpp{Dk5(
z!5PKLYFuwijraD2(Pd`1IU`<Z&F!|dHoWrw&}!Lt@9;{MZ?&x5hqF4)5hxb)c@z2q
zM_;74L~vPCwlcOc=SV_J#lk6EiI%&f6@fp#oWIQewa}N;ESbzRRNmJiI(`12j;jRf
zI;%>qqUd&Y3J&(jRmLDgoYq4TCWsVGO=vW!QS|1fCd6gV@R%1wVk$o~DLOM<Q5`XY
zYZk_J3!{%S`YGZB31u5=C>@PO==$m;M*x?yv3NX))c@&9O>2w8t1#5BQhQ}>S_dW2
zWI=Z57?5o`Zg9*X#Sp<wQ@$bGGUaPr+lW@fEr}`fBC-Ya3BvS(AY|Ap2pHBeLU5a7
z?iiR1$K0j3M=+YhcKZ(7ebby8*)|k^B#Ei~Sh?Mq3FtZ=5InTZY0NUGah~}R#bbhr
zg7NW0EFBKU;)WTMCP;(DaHjHmcRHMzmMI-i2&OHN84Khohh!<95j-!L5sAgszTleW
mFHDqri{(slN>fh9OM+PoWzIr*#VM~T-VnUiREl@qwf_LOBVcR*

delta 1166
zcmZ9LOH<QO9L3W#jxVZ$8#l~!BU4>4P}2fBV_Wcn1QahKRn!`wfhx2;?G1v8Rxt`H
zKB!us<G8~eBa3cXv0z~DCveXPaK@Y7w%nLxk&{XOzkANPnfqC{wN$r!Lf+M)HU*oT
zTh)Are7GuB((*4CSN@=yKdLnBE4{}ATzaFFThW!k+H8Kkm|u7OKF;4frm6+hx*Mvh
z_+y7mBMoMBk0eRM)QF|^j6RV?Xa-?lbT@lA5;s!AW;SLbQs~|9UgMvtezoB3tSrY?
zJ)Ls>$#x4J1gA8~rzx7tJ}FN7_?>-!W`69@5rVT6odo9`^C!_I%nvm2`J*LIOxd0k
zwj6%sLYo^k;mwVWkYW4@L9@_J(Br_K7qAyNwwK}}0f~CpkcurmIXWIkpNMW1tp%9+
z+a=dWE?MX&xa?pI2#hFaT%ov1Fla{iGKd|hkuo!hcm~%BgPL4UFBlYSV5&IKEqlU|
z>lSVh#3~H0=B2pl>yZ?8X<j=(ol}M=h6zTht))zwHYwtT$U$Y3H`L;&nyKQ_g6zrl
z8x}?h#vF`<gK>*9Zc~gCOhjv#k?dH)pq0ImoI%PL-C2D+g|rY~Yk6n%m6~JfzvjY{
zw9K08UX`TPRlq`qAgf85Vu5Nd3nr&eQrsb!5-`)47BFq)4Z_XZN;}}LSMOT5M{u9d
zo3ZCT;PW0*<Om*_f}L_`kA=|MSliqG2UErGwQf(Yeb&Mg0?QHFQ%7jeIA)Gwp5S?@
z{bVYUjmHv6yO|f;5Py@{zO9N|iwx57(!v74D+lDY1M-GL-cl?Q<V#kJL{cA*m94xJ
lD9vL3ex^93<-LUu1O*4>ql5B^Q$ACCA^2+Q6yJs#{{b6~POJa`

diff --git a/tests/data/linear_simulator_data_cube.npy b/tests/data/linear_simulator_data_cube.npy
index ecbeb984cb59498684c855be3e357d1eabd1be0d..8235bbeae41c6a2ea3e9da3e215e62754a16713e 100644
GIT binary patch
delta 1335
zcmZvb&rj1}7{|M=#HdVu1QQP~?BHQ~SO#>kY1A=97zLCU(fR{t85^69ZQt!Jf{IR|
zA`1SRIu#X(UPwHcaM)2#nsD^L@nUQ#kai{A;ms%O^Xc<_o_BV&erdJ-yN^BQ_4#})
zP0jv5z~AEM!HRZ<VWL?XcHvMz1vlU+u)|=)8WlI0|GbuEzuEQ~SZgz1Ji@se%m_Bg
z!xYve&i8AkxLz)<+x{FizU6P_OFv%PT@|dnQpL`-9~Y}saXu?WV>l^t>NyvaQV7c8
z9_>46Jd%}i(s&;2(+Ii6-6|Z7#WPVkACb{f>hEHI8tVplzI3pgZLDPBZq|9?l!8uz
z(*olbIKiep6kTrpq@BCx69*0xgeiIm1e0q5y%txC*Xq*Pa+r07`V{mNoH4o18e9Xq
zi%<*_oD&^}zNGY6DuyA8uW5(gH1@ZLb%xI?xIiG9d>0MAOS<ne#T9~KS=_ssZagFB
zlCd1FChFXY`p~~%>2t09OVP#dH%8dzw~kO{<-uOo5p0hrxJDptxm|*bV#M9YaOzKR
zX~{-)`zS?>Aih1?j7+mK#aL-&f&FdrsWDcrOq-W4e_do9g`tFkB*Ar)?}o{j(tYC;
zX@ZPctu~uaB;#~znX=P>TimscCy}$n-P~*)uf}XOcBdz=fD*_8BX9~P1e=1Jx^j|Y
zir|)^_7rYgYJ1fyp{;4P;U;-YowfEx{*Hoag1dU<ds^j!Uim)71A-aZ(9LFkvzGjP
z>RMFhr?T==o!wCgDGDAE%$f3gWXf+|S3ahALa?y0Ea^-#AB!Z@+Q1eqLVkxkuCj8l
zfpZi(o+@}o@Z2PPVG_R7g++>21g|$*j3?7lDze$)n@VHAtB#_gMUCmkwvvLk1n*47
V_a@_#Zu~&;k>HaoQ7n&m{sT0GcZC1|

delta 1308
zcmZXUO;giQ6o&KVI4T817Hm4!8ONzEYML}@&>2g?4<vwc6)6gq0*w@*?P+fiRI~~a
zQR@e&Rndh@W?VS3=)#RFonc_(kMMVx^wQc}bF;YfPI8{}zV{?+>rIR6O<xu6WI&YV
zR#BB@SqVxSa1owin6b18yneVBgV5*8al^nw8f@85KfICVzOePS77*<vNs^|CtH%US
zGc5rd^D()Nn4tVzSlukHZnD2mIB%7(ieJ%0t&q=io*J$Ac^9`7K4M1pIt^hXnvP{+
z(^&}3Lz_3U#}>!qMrzE=MokD8-1YpAN&#8*w<=P>-^d@V)kZow_o;RZIuM-J8Luwr
ztg|qj@sgdhKf659aumTi44nwhQ>+>2qFJgaOMbP)8Z4V4n&Vrr+ay;mw7pf=w!O6#
zGK@bV&@Jdj&_lJ*Th+n^(!xayeF#v2?cxrwmukp<RSRgeOO<GiI_C~wwxA!u6$&v>
zg@_QuAcm_5u9=a2cB4+tM#{`2;u#n!Bz5jzsEVqo0kP(!N?LKUn{%~=hb_2{AX?^m
zbPt9RZx18bx9)LrjS=D~hA{-=m4i*0IBjBx7uXg4SLMJIRSEi~TArg7zvnquuHUfW
z27(ESm!No)gm)9e6oTnUooz5XkuY$1l#=t1@<w)7<XK455z9fp^4}m4t)+o?`@n(>
zf~?Nyf(2M-Eiegk2E#1`vsJa{V6IfVt~?})XuCV!$RDDWA$w>8w=KAX;4Y~<@6^3V
z>fXnYL-4??^0JfO!<zJ@fNXb8e^(`~_-#M$%C#?8u!z8-(tAXuw?vSSF)Sl^QW}qB
zDv^yx6G`W|p3=ZTFxV=qG>~Y;fo8##>v(2C9>H@8_<{nyB*0e~RuHU~EE<VqY%02A
q@ik3UMMaisejP+3#NZnX-XbVa#CH_&Jt2O;u!i8H8N={t)b|$|R%Z$T


From 609625524f552d4f9db55a57d2161ba391e71e61 Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 14:39:45 -0400
Subject: [PATCH 09/15] Include space parameter in projector_data.npy

---
 tests/data/projector_data.npy | Bin 149024 -> 149043 bytes
 tests/test_projector.py       |   5 ++---
 2 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/tests/data/projector_data.npy b/tests/data/projector_data.npy
index f437e0e37b2af248869bb96944928c9a26f792d5..c36c71785a34e03a4bf6ae99b2bdff3b877074f8 100644
GIT binary patch
delta 223
zcmZ3`!@0SKbAu_9wSlRrsj;D%rIDebnT0_iQv@>u14D8_X(6+>ieE_~OQ2o@Q*0qa
z1V2#N5LNN!0;V;Ld@wOX3v)9=v&mX43L>aNg^Y=M3=NwTSq^@0mjAO|{tu%Yrwx)4
zb8|CfYke66aM}yB1E_lYLRO}72k8h_pp%LV5|dL4jUs?dhC<`g(!`QNlO#O=rXn^y

delta 204
zcmdno!?~b`bAu_9wTXd|xsj!*p`nGjrKwRNQv@>u14D8_X(6+>ieE_~OQ2o@Q*0qa
z1V2#N5LNN!0;V;Ld<ZcMV*``PS}Y18s6vH|iFynTn-f_Mes7llvt9lVqZ_9UNQtq5
kfu$u-kC_RwwZ04jSnWlq-oB8Psoa61v^24#&?rd{0NFD!`~Uy|

diff --git a/tests/test_projector.py b/tests/test_projector.py
index 2bbb7f8f..acaa447f 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -41,7 +41,6 @@ def init_data(path):
         config_dict = saved_data["config_dict"]
     else:
         config_dict = {}
-    config_dict["space"] = "real"
     config = AttrDict(config_dict)
     return saved_data, config
 
@@ -70,7 +69,7 @@ def test_projector_real():
     path = "tests/data/projector_data.npy"
 
     saved_data, config = init_data(path)
-    config.space = "real"
+    config["space"] = "real"
     rot_params = saved_data["rot_params"]
     projector = Projector(config)
     projector.vol = saved_data["volume"]
@@ -88,7 +87,7 @@ def test_projector_fourier():
     path = "tests/data/projector_data.npy"
 
     saved_data, config = init_data(path)
-    config.space = "fourier"
+    config["space"] = "fourier"
     rot_params = saved_data["rot_params"]
     projector = Projector(config)
     projector.vol = torch.fft.fftshift(

From d8de6466b9873d23e11893fe6ae6ab69c5a7000f Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 14:41:39 -0400
Subject: [PATCH 10/15] style fix

---
 simSPI/linear_simulator/projector.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index b83aa252..4cefb527 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -111,7 +111,7 @@ def _forward_fourier(self, rot_params):
 
         # rescale the coordinates to be compatible with the edge alignment of
         # torch.nn.functional.grid_sample
-        if 0 == self.config.side_len % 2:  # even case
+        if self.config.side_len % 2 == 0:  # even case
             rot_vol_coords = (
                 (rot_vol_coords + 1)
                 * (self.config.side_len)

From f5119453057bcedf1cfb77dd8b784ec8fb8645e5 Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 14:45:37 -0400
Subject: [PATCH 11/15] Some error checking

---
 simSPI/linear_simulator/projector.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index 4cefb527..502effd6 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -51,6 +51,10 @@ def __init__(self, config):
             # torch.nn.functional.grid_sample
             coords = 2 * coords
             self.register_buffer("vol_coords", coords.reshape(-1, 2))
+        else:
+            raise NotImplementedError(
+                f"Space type '{self.space}' " f"has not been implemented!"
+            )
 
     def forward(self, rot_params, proj_axis=-1):
         """Forward method for projection.
@@ -62,7 +66,13 @@ def forward(self, rot_params, proj_axis=-1):
         if self.space == "real":
             return self._forward_real(rot_params, proj_axis)
         elif self.space == "fourier":
+            if proj_axis != -1:
+                raise NotImplementedError("proj_axis must currently be -1 for Fourier space projection")
             return self._forward_fourier(rot_params)
+        else:
+            raise NotImplementedError(
+                f"Space type '{self.space}' " f"has not been implemented!"
+            )
 
     def _forward_fourier(self, rot_params):
         """Output the tomographic projection of the volume in Fourier space.

From e8f57ea056bab3194a2a8265a2a11ee284772254 Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 14:48:11 -0400
Subject: [PATCH 12/15] Remove elif after return

---
 simSPI/linear_simulator/projector.py | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index 502effd6..337ac1ae 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -65,14 +65,14 @@ def forward(self, rot_params, proj_axis=-1):
         """
         if self.space == "real":
             return self._forward_real(rot_params, proj_axis)
-        elif self.space == "fourier":
+        
+        if self.space == "fourier":
             if proj_axis != -1:
                 raise NotImplementedError("proj_axis must currently be -1 for Fourier space projection")
             return self._forward_fourier(rot_params)
-        else:
-            raise NotImplementedError(
-                f"Space type '{self.space}' " f"has not been implemented!"
-            )
+        raise NotImplementedError(
+            f"Space type '{self.space}' " f"has not been implemented!"
+        )
 
     def _forward_fourier(self, rot_params):
         """Output the tomographic projection of the volume in Fourier space.

From 6e1e537789edae9aebe7c43f469fea4c8111e00c Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 14:49:03 -0400
Subject: [PATCH 13/15] Shorten line

---
 simSPI/linear_simulator/projector.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index 337ac1ae..d5356d93 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -68,7 +68,9 @@ def forward(self, rot_params, proj_axis=-1):
         
         if self.space == "fourier":
             if proj_axis != -1:
-                raise NotImplementedError("proj_axis must currently be -1 for Fourier space projection")
+                raise NotImplementedError(
+                    "proj_axis must currently be -1 for Fourier space projection"
+                )
             return self._forward_fourier(rot_params)
         raise NotImplementedError(
             f"Space type '{self.space}' " f"has not been implemented!"

From eb6df3e742c125d63987ad62481d134c56491f93 Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 14:50:44 -0400
Subject: [PATCH 14/15] Remove commented out code

---
 simSPI/linear_simulator/projector.py | 4 ----
 tests/test_projector.py              | 5 -----
 2 files changed, 9 deletions(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index d5356d93..31a9ffff 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -114,12 +114,8 @@ def _forward_fourier(self, rot_params):
         rotmat = rot_params["rotmat"]
         batch_sz = rotmat.shape[0]
 
-        # print(rotmat[0])
         rotmat = torch.transpose(rotmat, -1, -2)
-        # print(rotmat[0])
         rot_vol_coords = self.vol_coords.repeat((batch_sz, 1, 1)).bmm(rotmat[:, :2, :])
-        # print(rot_vol_coords[0])
-        # print(rot_vol_coords[1])
 
         # rescale the coordinates to be compatible with the edge alignment of
         # torch.nn.functional.grid_sample
diff --git a/tests/test_projector.py b/tests/test_projector.py
index acaa447f..a80ef3a6 100644
--- a/tests/test_projector.py
+++ b/tests/test_projector.py
@@ -107,8 +107,3 @@ def test_projector_fourier():
     print("ratio", sz, 1 / (fft_proj_out.real[0, 0, 0, 0] / out.real[0, 0, 0, 0]))
     print("ratio", sz, 1 / (fft_proj_out.real[:, 0, 0, 0] / out.real[:, 0, 0, 0]))
     assert 0.01 > (fft_proj_out.real[0, 0, 0, 0] / out.real[0, 0, 0, 0] - 1).abs()
-    # print(out.shape[0],np.sqrt(out.shape[0]),1.0/np.sqrt(out.shape[0]))
-    # error_r = normalized_mse(fft_proj_out.real, out.real).item()
-    # error_i = normalized_mse(fft_proj_out.imag, out.imag).item()
-    # assert (error_r < 0.01) == 1
-    # assert (error_i < 0.01) == 1

From ffea6abd2b9134c6af6727e2d922c71a77c64752 Mon Sep 17 00:00:00 2001
From: Marcus Brubaker <mab@eecs.yorku.ca>
Date: Wed, 11 May 2022 14:55:55 -0400
Subject: [PATCH 15/15] Remove whitepsace

---
 simSPI/linear_simulator/projector.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/simSPI/linear_simulator/projector.py b/simSPI/linear_simulator/projector.py
index 31a9ffff..c9458b9f 100644
--- a/simSPI/linear_simulator/projector.py
+++ b/simSPI/linear_simulator/projector.py
@@ -65,7 +65,7 @@ def forward(self, rot_params, proj_axis=-1):
         """
         if self.space == "real":
             return self._forward_real(rot_params, proj_axis)
-        
+
         if self.space == "fourier":
             if proj_axis != -1:
                 raise NotImplementedError(