Fix CI (#411)

* Upgrade to nbdev3

* Clean and export

* Update test.yaml

* Fix pyvista dependency

* Update test.yaml

Author: eigenvivek

.github/workflows/test.yaml

@@ -23,7 +23,7 @@ jobs:
       - name: Doing editable install
         shell: bash -l {0}
         run: |
-          test -f setup.py && pip install -e ".[dev]"
+          pip install -e ".[dev]"
 
       - name: Check we are starting with clean git checkout
         shell: bash -l {0}
@@ -37,27 +37,27 @@ jobs:
       - name: Trying to strip out notebooks
         shell: bash -l {0}
         run: |
-          nbdev_clean
+          nbdev-clean
           git status -s # display the status to see which nbs need cleaning up
           if [[ `git status --porcelain -uno` ]]; then
             git status -uno
-            echo -e "!!! Detected unstripped out notebooks\n!!!Remember to run nbdev_install_hooks"
+            echo -e "!!! Detected unstripped out notebooks\n!!!Remember to run nbdev-install_hooks"
             echo -e "This error can also happen if you are using an older version of nbdev relative to what is in CI.  Please try to upgrade nbdev with the command `pip install -U nbdev`"
             false
           fi
 
-      - name: Run nbdev_export
+      - name: Run nbdev-export
         shell: bash -l {0}
         run: |
-          nbdev_export
+          nbdev-export
           if [[ `git status --porcelain -uno` ]]; then
-            echo "::error::Notebooks and library are not in sync.  Please run nbdev_export."
+            echo "::error::Notebooks and library are not in sync.  Please run nbdev-export."
             git status -uno
             git diff
             exit 1;
           fi
 
-      - name: Run nbdev_test
+      - name: Run nbdev-test
         shell: bash -l {0}
         run: |
-          nbdev_test
+          nbdev-test

diffdrr/data.py

@@ -1,6 +1,6 @@
 # AUTOGENERATED! DO NOT EDIT! File to edit: ../notebooks/api/03_data.ipynb.
 
-# %% ../notebooks/api/03_data.ipynb 3
+# %% ../notebooks/api/03_data.ipynb #cec0581f
 from __future__ import annotations
 
 from pathlib import Path
@@ -12,10 +12,10 @@
 from torchio import LabelMap, ScalarImage, Subject
 from torchio.transforms import Resample
 
-# %% auto 0
+# %% auto #0
 __all__ = ['load_example_ct', 'read']
 
-# %% ../notebooks/api/03_data.ipynb 5
+# %% ../notebooks/api/03_data.ipynb #126d05d3
 def load_example_ct(
     labels=None,
     orientation="AP",
@@ -37,7 +37,7 @@ def load_example_ct(
         **kwargs,
     )
 
-# %% ../notebooks/api/03_data.ipynb 6
+# %% ../notebooks/api/03_data.ipynb #9eda997b-6e88-4bb1-bb9f-313302fe3c1c
 from .pose import RigidTransform
 
 
@@ -96,7 +96,7 @@ def read(
             dtype=torch.float32,
         )
     elif orientation == "PA":
-        # Rotates the C-arm about the x-axis by 90 degrees
+        # Rotates the C-arm about the x-axis by 90 degrees 
         # Reverses the direction of the y-axis
         reorient = torch.tensor(
             [
@@ -180,7 +180,7 @@ def read(
 
     return subject
 
-# %% ../notebooks/api/03_data.ipynb 7
+# %% ../notebooks/api/03_data.ipynb #e20ad014-b2ae-41b9-8b65-e4ca865e19a2
 from diffdrr.pose import RigidTransform
 
 
@@ -210,7 +210,7 @@ def canonicalize(subject):
         subject.fiducials = affine(subject.fiducials)
     return subject
 
-# %% ../notebooks/api/03_data.ipynb 8
+# %% ../notebooks/api/03_data.ipynb #ba2941e0-cb0d-44c7-9b00-4dad1ced447d
 def transform_hu_to_density(volume, bone_attenuation_multiplier):
     # volume can be loaded as int16, need to convert to float32 to use float bone_attenuation_multiplier
     volume = volume.to(torch.float32)

diffdrr/detector.py

@@ -1,16 +1,16 @@
 # AUTOGENERATED! DO NOT EDIT! File to edit: ../notebooks/api/02_detector.ipynb.
 
-# %% ../notebooks/api/02_detector.ipynb 3
+# %% ../notebooks/api/02_detector.ipynb #b758e4f3
 from __future__ import annotations
 
 import torch
 from fastcore.basics import patch
 from torch.nn.functional import normalize
 
-# %% auto 0
+# %% auto #0
 __all__ = ['Detector', 'get_focal_length', 'get_principal_point', 'parse_intrinsic_matrix', 'make_intrinsic_matrix']
 
-# %% ../notebooks/api/02_detector.ipynb 5
+# %% ../notebooks/api/02_detector.ipynb #529b92a4-2f71-4d40-a25f-03cc4bc3eb6b
 from .pose import RigidTransform
 
 
@@ -93,7 +93,7 @@ def intrinsic(self):
         """The 3x3 intrinsic matrix."""
         return make_intrinsic_matrix(self).to(self.source)
 
-# %% ../notebooks/api/02_detector.ipynb 6
+# %% ../notebooks/api/02_detector.ipynb #b8ad63f4-0e38-4ea2-87b0-f298639dc9a9
 @patch
 def _initialize_carm(self: Detector):
     """Initialize the default position for the source and detector plane."""
@@ -137,7 +137,7 @@ def _initialize_carm(self: Detector):
         self.subsamples.append(sample.tolist())
     return source, target
 
-# %% ../notebooks/api/02_detector.ipynb 7
+# %% ../notebooks/api/02_detector.ipynb #063d06c3-2618-4282-accd-8fe0ab4d3faa
 from .pose import RigidTransform
 
 
@@ -153,7 +153,7 @@ def forward(self: Detector, extrinsic: RigidTransform, calibration: RigidTransfo
     target = pose(target)
     return source, target
 
-# %% ../notebooks/api/02_detector.ipynb 9
+# %% ../notebooks/api/02_detector.ipynb #4c0f02f6-c27e-4bdc-a204-31ba5c9f73de
 def get_focal_length(
     intrinsic,  # Intrinsic matrix (3 x 3 tensor)
     delx: float,  # X-direction spacing (in units length)
@@ -163,7 +163,7 @@ def get_focal_length(
     fy = intrinsic[1, 1]
     return abs((fx * delx) + (fy * dely)).item() / 2.0
 
-# %% ../notebooks/api/02_detector.ipynb 10
+# %% ../notebooks/api/02_detector.ipynb #a3535bdf-b819-4c42-9624-00d101b29ded
 def get_principal_point(
     intrinsic,  # Intrinsic matrix (3 x 3 tensor)
     height: int,  # Y-direction length (in units pixels)
@@ -175,7 +175,7 @@ def get_principal_point(
     y0 = dely * (intrinsic[1, 2] - height / 2)
     return x0.item(), y0.item()
 
-# %% ../notebooks/api/02_detector.ipynb 11
+# %% ../notebooks/api/02_detector.ipynb #750cb0fe-c96a-4c76-a2cd-51a74fdc6b05
 def parse_intrinsic_matrix(
     intrinsic,  # Intrinsic matrix (3 x 3 tensor)
     height: int,  # Y-direction length (in units pixels)
@@ -187,7 +187,7 @@ def parse_intrinsic_matrix(
     x0, y0 = get_principal_point(intrinsic, height, width, delx, dely)
     return focal_length, x0, y0
 
-# %% ../notebooks/api/02_detector.ipynb 12
+# %% ../notebooks/api/02_detector.ipynb #4e9f01cb-1dbc-4818-8521-e6785c101a82
 def make_intrinsic_matrix(detector: Detector):
     fx = detector.sdd / detector.delx
     fy = detector.sdd / detector.dely

diffdrr/drr.py

@@ -1,6 +1,6 @@
 # AUTOGENERATED! DO NOT EDIT! File to edit: ../notebooks/api/00_drr.ipynb.
 
-# %% ../notebooks/api/00_drr.ipynb 3
+# %% ../notebooks/api/00_drr.ipynb #ce95e1ac-413e-407b-87ad-0c7db3a945de
 from __future__ import annotations
 
 import numpy as np
@@ -11,10 +11,10 @@
 from .detector import Detector
 from .renderers import Siddon, Trilinear
 
-# %% auto 0
+# %% auto #0
 __all__ = ['DRR']
 
-# %% ../notebooks/api/00_drr.ipynb 7
+# %% ../notebooks/api/00_drr.ipynb #97297d06-6772-4dc7-8af5-d1ea7b379d8d
 from torchio import Subject
 
 from .pose import RigidTransform
@@ -138,15 +138,15 @@ def device(self):
     def dtype(self):
         return self.density.dtype
 
-# %% ../notebooks/api/00_drr.ipynb 8
+# %% ../notebooks/api/00_drr.ipynb #6513c593-32b8-4676-83d4-e9e7dcf0630a
 def reshape_subsampled_drr(img: torch.Tensor, detector: Detector, batch_size: int):
     n_points = detector.height * detector.width
     drr = torch.zeros(batch_size, n_points).to(img)
     drr[:, detector.subsamples[-1]] = img
     drr = drr.view(batch_size, 1, detector.height, detector.width)
     return drr
 
-# %% ../notebooks/api/00_drr.ipynb 10
+# %% ../notebooks/api/00_drr.ipynb #27b19dfc-6a15-4896-9faa-20faee84dc1f
 from torch.utils.checkpoint import checkpoint
 
 from .pose import RigidTransform, convert
@@ -168,12 +168,7 @@ def forward(
     if parameterization is None:
         pose = args[0]
     else:
-        pose = convert(
-            *args,
-            parameterization=parameterization,
-            convention=convention,
-            degrees=degrees,
-        )
+        pose = convert(*args, parameterization=parameterization, convention=convention, degrees=degrees)
 
     # Create the source / target points and render the image
     source, target = self.detector(pose, calibration)
@@ -231,7 +226,7 @@ def render(
 
     return img
 
-# %% ../notebooks/api/00_drr.ipynb 11
+# %% ../notebooks/api/00_drr.ipynb #d17edb1b-d3b4-4f31-b110-f5811ec6c183
 @patch
 def set_intrinsics_(
     self: DRR,
@@ -259,7 +254,7 @@ def set_intrinsics_(
         reverse_x_axis if reverse_x_axis is not None else self.detector.reverse_x_axis,
     ).to(self.density)
 
-# %% ../notebooks/api/00_drr.ipynb 12
+# %% ../notebooks/api/00_drr.ipynb #5d96d0d0-d76f-406e-8bc8-9023cf0f3e01
 @patch
 def rescale_detector_(self: DRR, scale: float):
     """Rescale the detector plane (inplace)."""
@@ -270,7 +265,7 @@ def rescale_detector_(self: DRR, scale: float):
         dely=float(self.detector.dely / scale),
     )
 
-# %% ../notebooks/api/00_drr.ipynb 13
+# %% ../notebooks/api/00_drr.ipynb #93a94ef3-5449-45dc-aa62-9fcf6fad643d
 @patch
 def perspective_projection(
     self: DRR,
@@ -294,7 +289,7 @@ def perspective_projection(
 
     return x[..., :2]
 
-# %% ../notebooks/api/00_drr.ipynb 14
+# %% ../notebooks/api/00_drr.ipynb #802ba874-eef8-4524-be5c-bd250e5639d7
 from torch.nn.functional import pad
 
 

diffdrr/metrics.py

@@ -1,23 +1,23 @@
 # AUTOGENERATED! DO NOT EDIT! File to edit: ../notebooks/api/05_metrics.ipynb.
 
-# %% ../notebooks/api/05_metrics.ipynb 3
+# %% ../notebooks/api/05_metrics.ipynb #86ff7dae
 from __future__ import annotations
 
 import torch
 
-# %% auto 0
+# %% auto #0
 __all__ = ['NormalizedCrossCorrelation2d', 'MultiscaleNormalizedCrossCorrelation2d', 'GradientNormalizedCrossCorrelation2d',
            'MutualInformation', 'LogGeodesicSE3', 'DoubleGeodesicSE3']
 
-# %% ../notebooks/api/05_metrics.ipynb 6
+# %% ../notebooks/api/05_metrics.ipynb #a77b3608-8d2a-43b6-b902-9f905877dd40
 from einops import rearrange
 
 
 def to_patches(x, patch_size):
     x = x.unfold(2, patch_size, step=1).unfold(3, patch_size, step=1).contiguous()
     return rearrange(x, "b c p1 p2 h w -> b (c p1 p2) h w")
 
-# %% ../notebooks/api/05_metrics.ipynb 7
+# %% ../notebooks/api/05_metrics.ipynb #28930479-d8e6-4859-b5de-38a5350f510b
 class NormalizedCrossCorrelation2d(torch.nn.Module):
     """Compute Normalized Cross Correlation between two batches of images."""
 
@@ -43,7 +43,7 @@ def norm(self, x):
         std = var.sqrt()
         return (x - mu) / std
 
-# %% ../notebooks/api/05_metrics.ipynb 8
+# %% ../notebooks/api/05_metrics.ipynb #8d06c00d-830c-48d9-b394-07cc83c1ed2b
 class MultiscaleNormalizedCrossCorrelation2d(torch.nn.Module):
     """Compute Normalized Cross Correlation between two batches of images at multiple scales."""
 
@@ -62,7 +62,7 @@ def forward(self, x1, x2):
             scores.append(weight * ncc(x1, x2))
         return torch.stack(scores, dim=0).sum(dim=0)
 
-# %% ../notebooks/api/05_metrics.ipynb 9
+# %% ../notebooks/api/05_metrics.ipynb #a6317e99-8a0a-4dce-959f-904c21595d71
 from torchvision.transforms.functional import gaussian_blur
 
 
@@ -92,7 +92,7 @@ def forward(self, img):
         x = self.filter(x)
         return x
 
-# %% ../notebooks/api/05_metrics.ipynb 10
+# %% ../notebooks/api/05_metrics.ipynb #fc39dd1d-ab40-4f7b-926d-dff305b9ab69
 class GradientNormalizedCrossCorrelation2d(NormalizedCrossCorrelation2d):
     """Compute Normalized Cross Correlation between the image gradients of two batches of images."""
 
@@ -103,7 +103,7 @@ def __init__(self, patch_size=None, sigma=1.0, **kwargs):
     def forward(self, x1, x2):
         return super().forward(self.sobel(x1), self.sobel(x2))
 
-# %% ../notebooks/api/05_metrics.ipynb 11
+# %% ../notebooks/api/05_metrics.ipynb #6e422fc1-8100-4120-b226-f6f54602fe3c
 from kornia.enhance.histogram import marginal_pdf, joint_pdf
 
 
@@ -118,7 +118,7 @@ def __init__(self, sigma=0.1, num_bins=256, epsilon=1e-10, normalize=True):
         self.normalize = normalize
 
     def forward(self, x1, x2):
-        assert x1.shape == x2.shape
+        assert(x1.shape == x2.shape)
         B, C, H, W = x1.shape
 
         x1 = x1.view(B, H * W, C)
@@ -138,7 +138,7 @@ def forward(self, x1, x2):
 
         return mutual_information
 
-# %% ../notebooks/api/05_metrics.ipynb 15
+# %% ../notebooks/api/05_metrics.ipynb #b691875b-c136-4ea5-8551-fab45530e315
 from .pose import RigidTransform, convert
 
 
@@ -157,7 +157,7 @@ def forward(
     ) -> Float[torch.Tensor, "b"]:
         return pose_2.compose(pose_1.inverse()).get_se3_log().norm(dim=1)
 
-# %% ../notebooks/api/05_metrics.ipynb 18
+# %% ../notebooks/api/05_metrics.ipynb #5ac6d6f4-462b-47ea-b25b-8a2518749e6f
 from .pose import so3_log_map
 
 
@@ -175,9 +175,7 @@ def __init__(
         self.sdr = sdd / 2
         self.eps = eps
 
-        self.rot_geo = lambda r1, r2: self.sdr * so3_log_map(
-            r1.transpose(-1, -2) @ r2
-        ).norm(dim=-1)
+        self.rot_geo = lambda r1, r2: self.sdr * so3_log_map(r1.transpose(-1, -2) @ r2).norm(dim=-1)
         self.xyz_geo = lambda t1, t2: (t1 - t2).norm(dim=-1)
 
     def forward(self, pose_1: RigidTransform, pose_2: RigidTransform):