search

point-cloud-radar / bird-cloud-gnn

This package generates graph representations from point cloud data and fits graph neural networks for classification problem
Apache License 2.0
8 stars 2 forks source link

Implement missing indicator features #124

Closed abelsiqueira closed 1 year ago

abelsiqueira commented 1 year ago

Description

Implement missing indicator features. The dataset.features include the new features with a _isna suffix.

Related issues:

Instructions to review the pull request

bart1 commented 1 year ago

I just see lint considers _processing_data too complex what do you think shall we leave it at this and refactor later when we know better what we want?

abelsiqueira commented 1 year ago

We ignore it, it's too stringent. But I will add a pylint ignore line to solve this in code.

sonarcloud[bot] commented 1 year ago

Kudos, SonarCloud Quality Gate passed!    Quality Gate passed

Bug A 0 Bugs
Vulnerability A 0 Vulnerabilities
Security Hotspot A 0 Security Hotspots
Code Smell A 0 Code Smells

95.8% 95.8% Coverage
0.0% 0.0% Duplication

bart1 commented 1 year ago

FYI I now have a test running to see what the effect is, each epoch is about 10 to 20% slower I think (for a 32 node, two layer model) so the performance hit is not enormous

abelsiqueira commented 1 year ago

That's a reasonable speed. Hopefully the gains are considerable.