Rules Extractor

WoodTapper rules extraction modules enables seamless integration of interpretable rule extraction into existing machine learning workflows. Here are an example on Iris data set.

Import modules

First, import necessary modules:

import pandas as pd
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.metrics import average_precision_score, accuracy_score,roc_auc_score

from woodtapper.extract_rules import SirusClassifier
from woodtapper.extract_rules.visualization import show_rules

Load data

iris = load_iris()
X = pd.DataFrame(iris.data, columns=iris.feature_names )
y = iris.target
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)

Train SirusClassifier

## RandomForestClassifier rules extraction
RFSirus = SirusClassifier(n_estimators=1000,max_depth=2,quantile=10,p0=0.0,max_n_rules=25, random_state=0,splitter="quantile")
RFSirus.fit(X_train,y_train)

Predictions

y_pred_sirus = RFSirus.predict(X_test)
y_pred_proba_sirus = RFSirus.predict_proba(X_test)

print('PR AUC :', average_precision_score(y_test, y_pred_proba_sirus))
print('ROC AUC :', roc_auc_score(y_test, y_pred_proba_sirus,average='micro',multi_class='ovr'))
print('Accuracy :', accuracy_score(y_test, y_pred_sirus))

Rules illustration

The rules are the same for all three classes but the output probabilities are specfici to each class:

show_rules(RFSirus,max_rules=10,target_class_index=0) ## show class Y=0

show_rules(RFSirus,max_rules=10,target_class_index=1) ## show class Y=1 through target_class_index=1 argument

show_rules(RFSirus,max_rules=10,target_class_index=2) ## show class Y=2