Changelog

Unreleased

Full Changelog

Closed issues:

to_graphviz() returns NoneType #139
RuntimeWarning: invalid value encountered in cast #137
np.float is deprecated #133
Failed to execute 'getPointAtLength' on 'SVGGeometryElement': The element's path is empty. #132
Error while running tree.render() #131
question about output #130
tree.render() throws error while working with data bricks azure #129
Prediction #128
Continuous Column name arg & Plotting Tree #127
Prior Nodes #126
Slow #125
missing value in ordinal feature and bonferroni adjustment #124
Feature selection difference with SPSS #123
how to get feature_importance #122
Issue exporting tree graph #121
Nominal Column is not defined #119
Documentation for library #118
I got the path to work, but now I'm back to the initial problem I had with the invalid argument in trees. -_- #117
Not being able to visualize it in Colab #116
is it copy of https://github.com/codingblg/CHAID\_phi #115

Merged pull requests:

Fix one numpy deprecated function (in1d -> isin) #140 (jihaekor)
Use an explicit sentinel value rather than relying on integer cast of np.nan #138 (jihaekor)
Add a new max_splits parameter; make a couple of fix updates #136 (jihaekor)
Fix errors due to numpy deprecation; split required packages further; update specs #135 (jihaekor)
Fixed README imports #120 (Rambatino)
Added exhaustive CHAID #113 (Rambatino)
Added graph specs and switched off true divide warnings for specs #111 (Rambatino)

v5.4.0 (2020-10-30)

Full Changelog

Closed issues:

Exhaustive chaid #112
Not being able to visualize the tree #110
No Attribute "from_numpy" #109
Error while importing in Jupyter Notebook #108
Issue with members property #107
Need to spec out the graphing architecture #106
Warnings have appeared when running specs locally. The bit rot is real #105
Issue in running the "tree.render(path=None, view=False)" #99

v5 (2020-01-13)

Full Changelog

v5.3.0 (2020-01-13)

Full Changelog

Closed issues:

Missing dependencies #102
Python 2 incompatibility: invalid_split_reason.py uses enum #101
tree.to_tree() is not working for creating the dot file in CHAID analysis and also tried "tree.render(path=None, view=False)" to get tree but failed get so.. please any suggestion #98
Export error #97
how to graphically display the tree? #93

Merged pull requests:

Fix graph installations in the README.md. Add note about python 2.7 #104 (Rambatino)
Py2 compat: add conditional dep #103 (mjpieters)
Refactor Graph to avoid cross-platform issues #100 (mjpieters)

v5.2.0 (2019-04-03)

Full Changelog

Implemented enhancements:

Added render method and graph export with path variable #94 (Rambatino)
Added accuracy method #90 (Rambatino)

Closed issues:

IS ID5 tree equalst to CHAID in concepts? #92
Make predictions on testing set and calculate the propensity scores #89
Risk broken, needs speccing #74

Merged pull requests:

Attempting new cirlce ci file #95 (Rambatino)

v5.1.0 (2018-10-03)

Full Changelog

v5.0.5 (2018-09-19)

Full Changelog

Fixed bugs:

Created passing unit test case for issue in #87 #88 (Rambatino)

Closed issues:

model_predictions fails with categorical dependant variables #87
Couple questions #85
Why isn't there a predict function ? #72

Merged pull requests:

House keeping #86 (Rambatino)
Reverted back heuristic approach #84 (Rambatino)

v5.0.4 (2017-12-20)

Full Changelog

v5.0.3 (2017-09-29)

Full Changelog

Fixed bugs:

Valid splits discounted when most significant split is generated below base size #81
Fixing minimum node sizes whereby combined choices aren't selected #83 (Rambatino)

Closed issues:

Maybe an error in the doc? #79

v5.0.2 (2017-08-24)

Full Changelog

Fixed bugs:

Specs and fix #78 (Rambatino)

v5.0.1 (2017-08-22)

Full Changelog

Fixed bugs:

Fixed numpy for 1.11.1 #77 (Rambatino)

v5.0.0 (2017-08-21)

Full Changelog

Implemented enhancements:

Formalised tree initialisation method to use Column classes #76 (Rambatino)

Closed issues:

Continuous independent variables? #75
CHAID tree to json #71
Output Tree as pandas DataFrame #69
Basic clarifications #66

Merged pull requests:

First draft of the how to guide #67 (Rambatino)

v4.0.0 (2017-06-14)

Full Changelog

Implemented enhancements:

Changed architecture of from_pandas_df to align variable types and instance variables into a single parameter #70 (Rambatino)
Adding invalid split messages #68 (Rambatino)