goodok / fastai_sparse Goto Github PK

print_versions returns None due to typo in the following line
https://github.com/goodok/fastai_sparse/blob/master/fastai_sparse/utils.py#L70

https://github.com/goodok/fastai_sparse/blob/master/fastai_sparse/transforms/main.py#L195

It looks like that random selection in _sample_point transform is implemented with replacement. Think it's better to use
numpy.random.choice with replace=False

Hi,

there is an issues with ipvolums or widgets:

In the /fastai_sparse/notebooks/transforms/transforms.ipynb notebook:

in the cell:
o.show(with_normals=False, point_size_value=0.0, labels=o.labels[0])

just text is displayed:
VBox(children=(VBox(children=(Figure(camera=PerspectiveCamera(fov=46.0, position=(0.0, 0.0,

These commands have been executed successfully
jupyter nbextension install --py widgetsnbextension --user
jupyter nbextension enable widgetsnbextension --user --py
jupyter nbextension install --py ipyvolume --user
jupyter nbextension enable ipyvolume --user --py

It seems that the transformation of color noise does not look like color noise at each point, but simply makes a random change over 3 channels common to all points

https://nbviewer.jupyter.org/github/goodok/fastai_sparse/blob/master/notebooks/transforms/transforms.ipynb

x_colors = pc.colors
amplitude = 0.1
x_colors_b = x_colors +
np.random.randn(3).astype(np.float32) * amplitude
x_colors - x_colors_b

returns differ between before and after:

array([[0.03873268, 0.03023028, 0.1048553 ],
[0.03873268, 0.03023027, 0.1048553 ],
[0.03873268, 0.03023028, 0.1048553 ],
...,
[0.03873268, 0.03023028, 0.1048553 ],
[0.03873265, 0.03023028, 0.1048553 ],
[0.03873265, 0.03023028, 0.1048553 ]], dtype=float32)