search

automl / HPOlib1.5

GNU General Public License v3.0
69 stars 27 forks source link

Unittests #39

Closed KEggensperger closed 7 years ago

KEggensperger commented 7 years ago

This PR adds two new unittests

1) check whether RNGs are used correctly 2) Evaluate random configurations (using pynisher to terminate long running cfgs)

codecov-io commented 7 years ago

Codecov Report

Merging #39 into development will increase coverage by 16.1%. The diff coverage is 41.81%.

Impacted file tree graph

@@               Coverage Diff               @@
##           development      #39      +/-   ##
===============================================
+ Coverage        55.93%   72.03%   +16.1%     
===============================================
  Files               26       26              
  Lines             1314     1334      +20     
===============================================
+ Hits               735      961     +226     
+ Misses             579      373     -206
Impacted Files Coverage Δ
...etic_functions/wrapper/synthetic_noise_and_cost.py 33.33% <ø> (ø)
hpolib/benchmarks/synthetic_functions/__init__.py 100% <100%> (ø) :arrow_up:
...polib/benchmarks/synthetic_functions/rosenbrock.py 94.28% <100%> (+42.85%) :arrow_up:
hpolib/abstract_benchmark.py 87.23% <100%> (+0.56%) :arrow_up:
hpolib/util/data_manager.py 98.23% <100%> (+25.45%) :arrow_up:
hpolib/benchmarks/ml/fully_connected_network.py 27.81% <18.18%> (+27.81%) :arrow_up:
hpolib/benchmarks/ml/logistic_regression.py 50.87% <20%> (+0.87%) :arrow_up:
hpolib/benchmarks/ml/conv_net.py 34.26% <22.22%> (+34.26%) :arrow_up:
hpolib/benchmarks/ml/svm_benchmark.py 56.75% <33.33%> (+56.75%) :arrow_up:
hpolib/benchmarks/ml/autosklearn_benchmark.py 95.29% <85.71%> (+0.17%) :arrow_up:
... and 9 more

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 bbf3e53...06997a6. Read the comment docs.