Closed dllu closed 1 year ago
The comma makes it a tuple (None, ) instead of just None. Then, it fails with:
(None, )
None
In [3]: scene = GaussianPointCloudScene.from_parquet('/home/danlu/debug/gaussian-splatting-2/logs/scene_169000.parquet') --------------------------------------------------------------------------- TypeError Traceback (most recent call last) Cell In[3], line 1 ----> 1 scene = GaussianPointCloudScene.from_parquet('/home/danlu/debug/gaussian-splatting-2/logs/scene_169000.parquet') File ~/builds/taichi_3d_gaussian_splatting/taichi_3d_gaussian_splatting/GaussianPointCloudScene.py:150, in GaussianPointCloudScene.from_parquet(path, config) 147 else: 148 valid_point_cloud_features = torch.from_numpy( 149 scene_df[feature_columns].to_numpy()) --> 150 scene = GaussianPointCloudScene( 151 point_cloud, config, point_cloud_features=valid_point_cloud_features) 152 return scene File ~/builds/taichi_3d_gaussian_splatting/taichi_3d_gaussian_splatting/GaussianPointCloudScene.py:42, in GaussianPointCloudScene.__init__(self, point_cloud, config, point_cloud_features) 40 if config.max_num_points_ratio is not None: 41 num_points = point_cloud.shape[0] ---> 42 max_num_points = int(num_points * config.max_num_points_ratio) 43 assert max_num_points > num_points, "max_num_points_ratio should be greater than 1.0" 44 point_cloud = torch.cat( 45 [point_cloud, torch.zeros((max_num_points - num_points, 3))], dim=0) TypeError: int() argument must be a string, a bytes-like object or a number, not 'tuple'
Thanks a lot for your contribution!
The comma makes it a tuple
(None, )
instead of justNone
. Then, it fails with: