[#399] removed fraudulent demand regex

[Samuel-Campbell]

• [x] Remove demand regex (bonus task)
• [x] Change additional_indemnity_money to boolean
• [x] Removed additional indemnity money regressor
• [x] Test pipeline from beginning to end (make sure nothing breaks) - using sample of 1000
• [x] Removed outcomes with low occurrence

1- Removing demand regex makes tagging quicker. They are not used anyways 2- Additional indemnity money will now become a boolean since we will be calculating it's value using a formula that we know of 3- Had to remove outcomes with low occurrence --> delcares housing inhabitable. These outcomes cannot be implemented without a custom multi class svm wrapper. Right now the default one from sklearn doesn't always work. Since the training and testing sets are randomly split AND the outcome occurrence is very low THEN there is a chance that there are 0 values for a given outcome in the training set. If this happens then no classifier is created. You will get null pointer errors from this in the future.

[codecov[bot]]

Codecov Report

Merging #413 into master will decrease coverage by 0.34%. The diff coverage is 50%.

@@            Coverage Diff             @@
##           master     #413      +/-   ##
==========================================
- Coverage   70.87%   70.53%   -0.35%     
==========================================
  Files          54       53       -1     
  Lines        2249     2182      -67     
  Branches       23       23              
==========================================
- Hits         1594     1539      -55     
+ Misses        654      642      -12     
  Partials        1        1

Impacted Files	Coverage Δ
...raction/post_processing/regex/regex_test_helper.py	53.06% <ø> (-2.3%)	:arrow_down:
...ion/post_processing/regex/misc/regex_lib_helper.py	28.12% <ø> (-1.01%)	:arrow_down:
...e_extraction/post_processing/regex/regex_tagger.py	98.18% <ø> (+2.08%)	:arrow_up:
...ture_extraction/post_processing/regex/regex_lib.py	100% <ø> (ø)	:arrow_up:
...regression/multi_output/multi_output_regression.py	48.48% <50%> (+1.26%)	:arrow_up:
...ion/single_output_regression/abstract_regressor.py	44.59% <0%> (-24.33%)	:arrow_down:

---

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 358df26...ec6d6aa. Read the comment docs.