[#43] fixed bug that prevents weights

[Samuel-Campbell]

1) Some predictors don't have enough data and the multiclass SVM creates a dummy predictor object. Try catch when this occurs and report an error with the outcome

[codecov[bot]]

Codecov Report

Merging #392 into master will increase coverage by 2.53%. The diff coverage is 90.9%.

@@            Coverage Diff             @@
##           master     #392      +/-   ##
==========================================
+ Coverage   65.67%   68.21%   +2.53%     
==========================================
  Files          25       51      +26     
  Lines        1005     2095    +1090     
  Branches       17       17              
==========================================
+ Hits          660     1429     +769     
- Misses        340      661     +321     
  Partials        5        5

Impacted Files	Coverage Δ
src/ml_service/init.py	74.07% <ø> (ø)
...raining/classifier/multi_output/multi_class_svm.py	86.89% <90.9%> (ø)
src/ml_service/web/ml_controller.py	63.23% <0%> (ø)
...ture_extraction/post_processing/regex/regex_lib.py	100% <0%> (ø)
...ion/single_output_regression/abstract_regressor.py	68.91% <0%> (ø)
...ion/post_processing/regex/misc/regex_lib_helper.py	29.12% <0%> (ø)
...ice/model_training/regression/regression_driver.py	60% <0%> (ø)
...ce/feature_extraction/feature_extraction_driver.py	100% <0%> (ø)
src/ml_service/util/constant.py	100% <0%> (ø)
...n/post_processing/regex/regex_entity_extraction.py	83.33% <0%> (ø)
... and 18 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 cc214f3...25a10eb. Read the comment docs.