datascienceinc / Skater / 454 / 3
62%
master: 62%

DEFAULT BRANCH: master
Ran
Files 57
Run time 27s
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 62.296% (+1.0%) from 61.337%
454.3

push

travis-ci

web-flow 
Add support for post pruning for Tree Surrogates (#253)

* initial skeleton for post pruning of trees

* corrections and improvement to post pruning algorithm and more logging

* minor correction in the logic for post pruning

* more logging, and update notebook example

* updated example with text-based visualization of decision stubs

* correction in resolving access to respective scorer and addition of more unit tests and show_in_notebook function

* correcting the use of name variable in the scorer and improving code styling

* minor corrections to tests

* code improvements

* updates to the logic of identifying node as leaf nodes for visualization

* support for adding style to iframe rendering for some simple interaction

* Updating the example notebook with recent changes to API and improvements

* replacing __name__ with name and adding main to allow running test independently

* better use of base model internally to compute artifacts as needed

* better use of logging and more code improvements

* updated example with recent API changes

* added simple tests for tree surrogates, some more code improvement all around

* removing missed unused import

* adding ability to handle either predicted values or probabilities while optimizing tree

* minor correction to function name

* another variable name change(typo correction)

* adjusting tests with new changes, and removing unused imports

* more changes related to tests

* added more tests and corrected pre-pruning to use a specified scoring function

* added scorer_type as a parameter to _pre_pruning

* improving logging for pre-pruning

* correcting assigning probability flag for regressor and updating examples

* renaming learn to fit, updated the tests as well

* updating the example with the renaming of 'learn' to 'fit'

* correcting a minor code error resulting in test failures

* more updates to doc-strings

* more documenta... (continued)

1834 of 2944 relevant lines covered (62.3%)

0.62 hits per line

Source Files on job 454.3