Notebook source code: notebooks/how_to/04_descriptor_pipeline.ipynb
Run it yourself on binder

How to create a descriptor pipeline?#

 In [1]:

from geomfum.dataset import NotebooksDataset
from geomfum.descriptor.pipeline import (
    ArangeSubsampler,
    DescriptorPipeline,
    L2InnerNormalizer,
)
from geomfum.descriptor.spectral import HeatKernelSignature, WaveKernelSignature
from geomfum.shape import TriangleMesh

Load a mesh.

 In [2]:

dataset = NotebooksDataset()
mesh = TriangleMesh.from_file(dataset.get_filename("cat-00"))

INFO: Data has already been downloaded... using cached file ('/home/ubuntu/.geomfum/data/cat-00.off').
/home/ubuntu/giulio_vigano/geomfum_proj/venv/lib/python3.12/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
  from .autonotebook import tqdm as notebook_tqdm

Set Laplace eigenbasis.

 In [3]:

mesh.laplacian.find_spectrum(spectrum_size=10, set_as_basis=True)

mesh.basis

 Out [3]:

<geomfum.basis.LaplaceEigenBasis at 0x7764d94f4740>

A pipeline is composed of `Descriptor <./03_descriptors.ipynb>`__, Subsampler(./sampler.ipynb) and Normalizer.

 In [4]:

steps = [
    HeatKernelSignature.from_registry(n_domain=4),
    ArangeSubsampler(subsample_step=2),
    WaveKernelSignature.from_registry(n_domain=3),
    L2InnerNormalizer(),
]

pipeline = DescriptorPipeline(steps)

 In [5]:

descr = pipeline.apply(mesh)

descr.shape

 Out [5]:

(5, 7207)

How to create a descriptor pipeline?#

Further reading#

This Page