- Install Hatch (this is a python environment manager)
- Clone repository
- Run
hatch shell
to enter the virtual environment - Put
collected_splats.csv
andgt.png
in thesrc/clustering_exploration/data/
directory- These were generated using this fork of
diff-gaussian-rasterization
- These were generated using this fork of
- Run through the notebooks in
src/cluster_exploration
.
- The test setup can be found in
tests
folder - Before running any program, modify the
config.json
with correct file path. dataset_init.py
will create csv files for views from given model.- Usage:
python dataset_init.py <path-to-config>
- Usage:
test_suite.py
will compute images and corresponding statistics from clustering algorithm.- Usage:
python test_suite.py <path-to-config>
- Usage:
clustering-exploration
is distributed under the terms of the MIT license.