format

gsplat/__init__.py

@@ -2,26 +2,29 @@
 import torch
 from .project_gaussians import project_gaussians
 from .rasterize import rasterize_gaussians
-from .utils import map_gaussian_to_intersects, bin_and_sort_gaussians, compute_cumulative_intersects, compute_cov2d_bounds, get_tile_bin_edges
+from .utils import (
+    map_gaussian_to_intersects,
+    bin_and_sort_gaussians,
+    compute_cumulative_intersects,
+    compute_cov2d_bounds,
+    get_tile_bin_edges,
+)
 from .sh import spherical_harmonics
 from .version import __version__
 import warnings
 
 
 __all__ = [
     "__version__",
-
     "project_gaussians",
     "rasterize_gaussians",
     "spherical_harmonics",
-
     # utils
     "bin_and_sort_gaussians",
     "compute_cumulative_intersects",
     "compute_cov2d_bounds",
     "get_tile_bin_edges",
     "map_gaussian_to_intersects",
-
     # Function.apply() will be deprecated
     "ProjectGaussians",
     "RasterizeGaussians",
@@ -36,102 +39,128 @@
 
 # Define these for backwards compatibility
 
-class MapGaussiansToIntersects(torch.autograd.Function):
 
+class MapGaussiansToIntersects(torch.autograd.Function):
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("MapGaussiansToIntersects is deprecated, use map_gaussian_to_intersects instead", DeprecationWarning)
+        warnings.warn(
+            "MapGaussiansToIntersects is deprecated, use map_gaussian_to_intersects instead",
+            DeprecationWarning,
+        )
         return map_gaussian_to_intersects(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
-
+
+
 class ComputeCumulativeIntersects(torch.autograd.Function):
-
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("ComputeCumulativeIntersects is deprecated, use compute_cumulative_intersects instead", DeprecationWarning)
+        warnings.warn(
+            "ComputeCumulativeIntersects is deprecated, use compute_cumulative_intersects instead",
+            DeprecationWarning,
+        )
         return compute_cumulative_intersects(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
-
+
+
 class ComputeCov2dBounds(torch.autograd.Function):
-
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("ComputeCov2dBounds is deprecated, use compute_cov2d_bounds instead", DeprecationWarning)
+        warnings.warn(
+            "ComputeCov2dBounds is deprecated, use compute_cov2d_bounds instead",
+            DeprecationWarning,
+        )
         return compute_cov2d_bounds(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
 
+
 class GetTileBinEdges(torch.autograd.Function):
-
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("GetTileBinEdges is deprecated, use get_tile_bin_edges instead", DeprecationWarning)
+        warnings.warn(
+            "GetTileBinEdges is deprecated, use get_tile_bin_edges instead",
+            DeprecationWarning,
+        )
         return get_tile_bin_edges(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
-
+
+
 class BinAndSortGaussians(torch.autograd.Function):
-
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("BinAndSortGaussians is deprecated, use bin_and_sort_gaussians instead", DeprecationWarning)
+        warnings.warn(
+            "BinAndSortGaussians is deprecated, use bin_and_sort_gaussians instead",
+            DeprecationWarning,
+        )
         return bin_and_sort_gaussians(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
-
+
+
 class ProjectGaussians(torch.autograd.Function):
-
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("ProjectGaussians is deprecated, use project_gaussians instead", DeprecationWarning)
+        warnings.warn(
+            "ProjectGaussians is deprecated, use project_gaussians instead",
+            DeprecationWarning,
+        )
         return project_gaussians(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
-
+
+
 class RasterizeGaussians(torch.autograd.Function):
-
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("RasterizeGaussians is deprecated, use rasterize_gaussians instead", DeprecationWarning)
+        warnings.warn(
+            "RasterizeGaussians is deprecated, use rasterize_gaussians instead",
+            DeprecationWarning,
+        )
         return rasterize_gaussians(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
-
+
+
 class NDRasterizeGaussians(torch.autograd.Function):
-
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("NDRasterizeGaussians is deprecated, use rasterize_gaussians instead", DeprecationWarning)
+        warnings.warn(
+            "NDRasterizeGaussians is deprecated, use rasterize_gaussians instead",
+            DeprecationWarning,
+        )
         return rasterize_gaussians(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
 
-class SphericalHarmonics(torch.autograd.Function):
 
+class SphericalHarmonics(torch.autograd.Function):
     @staticmethod
     def forward(ctx, *args, **kwargs):
-        warnings.warn("SphericalHarmonics is deprecated, use spherical_harmonics instead", DeprecationWarning)
+        warnings.warn(
+            "SphericalHarmonics is deprecated, use spherical_harmonics instead",
+            DeprecationWarning,
+        )
         return spherical_harmonics(*args, **kwargs)
-    
+
     @staticmethod
     def backward(ctx: Any, *grad_outputs: Any) -> Any:
         raise NotImplementedError
-

gsplat/project_gaussians.py

@@ -8,6 +8,7 @@
 
 import gsplat.cuda as _C
 
+
 def project_gaussians(
     means3d: Float[Tensor, "*batch 3"],
     scales: Float[Tensor, "*batch 3"],
@@ -28,7 +29,7 @@ def project_gaussians(
 
     Note:
         This function is differentiable w.r.t the means3d, scales and quats inputs.
-    
+
     Args:
        means3d (Tensor): xyzs of gaussians.
        scales (Tensor): scales of the gaussians.
@@ -210,4 +211,4 @@ def backward(ctx, v_xys, v_depths, v_radii, v_conics, v_num_tiles_hit, v_cov3d):
             None,
             # clip_thresh,
             None,
-        )
+        )

gsplat/rasterize.py

@@ -54,7 +54,9 @@ def rasterize_gaussians(
             background.shape[0] == colors.shape[-1]
         ), f"incorrect shape of background color tensor, expected shape {colors.shape[-1]}"
     else:
-        background = torch.ones(colors.shape[-1], dtype=torch.float32, device=colors.device)
+        background = torch.ones(
+            colors.shape[-1], dtype=torch.float32, device=colors.device
+        )
 
     if xys.ndimension() != 2 or xys.size(1) != 2:
         raise ValueError("xys must have dimensions (N, 2)")
@@ -103,9 +105,7 @@ def forward(
         block = (BLOCK_X, BLOCK_Y, 1)
         img_size = (img_width, img_height, 1)
 
-        num_intersects, cum_tiles_hit = compute_cumulative_intersects(
-            num_tiles_hit
-        )
+        num_intersects, cum_tiles_hit = compute_cumulative_intersects(num_tiles_hit)
 
         (
             isect_ids_unsorted,
@@ -198,4 +198,3 @@ def backward(ctx, v_out_img):
             None,  # img_width
             None,  # background
         )
-

gsplat/sh.py

@@ -32,6 +32,7 @@ def deg_from_sh(num_bases: int):
         return 4
     assert False, "Invalid number of SH bases"
 
+
 def spherical_harmonics(
     degrees_to_use: int,
     viewdirs: Float[Tensor, "*batch 3"],
@@ -46,12 +47,15 @@ def spherical_harmonics(
         degrees_to_use (int): degree of SHs to use (<= total number available).
         viewdirs (Tensor): viewing directions.
         coeffs (Tensor): harmonic coefficients.
-    
+
     Returns:
         The spherical harmonics.
     """
     assert coeffs.shape[-2] >= num_sh_bases(degrees_to_use)
-    return _SphericalHarmonics.apply(degrees_to_use, viewdirs.contiguous(), coeffs.contiguous())
+    return _SphericalHarmonics.apply(
+        degrees_to_use, viewdirs.contiguous(), coeffs.contiguous()
+    )
+
 
 class _SphericalHarmonics(Function):
     """Compute spherical harmonics
@@ -91,4 +95,3 @@ def backward(ctx, v_colors: Float[Tensor, "*batch 3"]):
                 num_points, degree, degrees_to_use, viewdirs, v_colors
             ),
         )
-

gsplat/utils.py

@@ -39,10 +39,17 @@ def map_gaussian_to_intersects(
         - **gaussian_ids** (Tensor): Tensor that maps isect_ids back to cum_tiles_hit.
     """
     isect_ids, gaussian_ids = _C.map_gaussian_to_intersects(
-        num_points, num_intersects, xys.contiguous(), depths.contiguous(), radii.contiguous(), cum_tiles_hit.contiguous(), tile_bounds
+        num_points,
+        num_intersects,
+        xys.contiguous(),
+        depths.contiguous(),
+        radii.contiguous(),
+        cum_tiles_hit.contiguous(),
+        tile_bounds,
     )
     return (isect_ids, gaussian_ids)
 
+
 def get_tile_bin_edges(
     num_intersects: int, isect_ids_sorted: Int[Tensor, "num_intersects 1"]
 ) -> Int[Tensor, "num_intersects 2"]:
@@ -96,7 +103,7 @@ def compute_cumulative_intersects(
 
     Note:
         This function is not differentiable to any input.
-    
+
     Args:
         num_tiles_hit (Tensor): number of intersected tiles per gaussian.
 
@@ -128,7 +135,7 @@ def bin_and_sort_gaussians(
 ]:
     """Mapping gaussians to sorted unique intersection IDs and tile bins used for fast rasterization.
 
-    We return both sorted and unsorted versions of intersect IDs and gaussian IDs for testing purposes. 
+    We return both sorted and unsorted versions of intersect IDs and gaussian IDs for testing purposes.
 
     Note:
         This function is not differentiable to any input.