Speed-up tests of pipeline combinations using `subsample` with `random_state`

GlacioHack / xdem

Analysis of digital elevation models (DEMs)

https://xdem.readthedocs.io/

MIT License

139 stars 39 forks source link

Speed-up tests of pipeline combinations using `subsample` with `random_state` #538

Closed rhugonnet closed 4 months ago

rhugonnet commented 4 months ago

Those currently take a while because they use the full sample, and (very rarely) randomly fail because of DirectionalBias: could use another fit_func than the sum of sinusoids to avoid this.