search

traja-team / traja

Python tools for spatial trajectory and time-series data analysis
https://traja.readthedocs.io
MIT License
98 stars 25 forks source link

added validation and testing during training #72

Closed MaddyThakker closed 3 years ago

MaddyThakker commented 3 years ago

added params to perform evaluation on validation and test datasets to get an idea of how the model performs on those datasets while training.

codecov[bot] commented 3 years ago

Codecov Report

Merging #72 (cba433d) into master (f87978d) will increase coverage by 0.72%. The diff coverage is 96.96%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master      #72      +/-   ##
==========================================
+ Coverage   75.66%   76.39%   +0.72%     
==========================================
  Files          24       24              
  Lines        2466     2482      +16     
==========================================
+ Hits         1866     1896      +30     
+ Misses        600      586      -14
Impacted Files Coverage Δ
traja/models/train.py 97.09% <96.96%> (+9.27%) :arrow_up:

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update f87978d...cba433d. Read the comment docs.

WolfByttner commented 3 years ago

There are several lines that are not being tested. Would it be possible to write a test that showcases the new functionality?

MaddyThakker commented 3 years ago

There are several lines that are not being tested. Would it be possible to write a test that showcases the new functionality?

@WolfByttner Done. Review again. Thanks.

WolfByttner commented 3 years ago

validate_every=None and test_every=None are not tested. I know it isn't much, but I don't want functionality like that to regress at some later stage.