And is there not also the case where we play and--make up the rules as we go along?
-Ludwig Wittgenstein
This package implements two interpretable coverage-based ruleset algorithms: IREP and RIPPERk, as well as additional features for model interpretation.
Performance is similar to sklearn's DecisionTree CART implementation (see Performance Tests).
For explanation of the algorithms, see my article in Towards Data Science, or the papers below, under Useful References.
To install, use
$ pip install wittgensteinTo uninstall, use
$ pip uninstall wittgenstein- pandas
- numpy
- python version>=3.6
Usage syntax is similar to sklearn's.
Once you have loaded and split your data...
>>> import pandas as pd
>>> df = pd.read_csv(dataset_filename)
>>> from sklearn.model_selection import train_test_split # Or any other mechanism you want to use for data partitioning
>>> train, test = train_test_split(df, test_size=.33)Use the fit method to train a RIPPER or IREP classifier:
>>> import wittgenstein as lw
>>> ripper_clf = lw.RIPPER() # Or irep_clf = lw.IREP() to build a model using IREP
>>> ripper_clf.fit(df, class_feat='Poisonous/Edible', pos_class='p') # Or pass X and y data to .fit
>>> ripper_clf
<RIPPER(max_rules=None, random_state=2, max_rule_conds=None, verbosity=0, max_total_conds=None, k=2, prune_size=0.33, dl_allowance=64, n_discretize_bins=10) with fit ruleset> # Hyperparameter details available in the docstrings and TDS article belowAccess the underlying trained model with the ruleset_ attribute, or output it with out_model(). A ruleset is a disjunction of conjunctions -- 'V' represents 'or'; '^' represents 'and'.
In other words, the model predicts positive class if any of the inner-nested condition-combinations are all true:
>>> ripper_clf.out_model() # or ripper_clf.ruleset_
[[Odor=f] V
[Gill-size=n ^ Gill-color=b] V
[Gill-size=n ^ Odor=p] V
[Odor=c] V
[Spore-print-color=r] V
[Stalk-surface-below-ring=y ^ Stalk-surface-above-ring=k] V
[Habitat=l ^ Cap-color=w] V
[Stalk-color-above-ring=y]]IREP models tend be higher bias, RIPPER's higher variance.
To score a trained model, use the score function:
>>> X_test = test.drop('Poisonous/Edible', axis=1)
>>> y_test = test['Poisonous/Edible']
>>> ripper_clf.score(test_X, test_y)
0.9985686906328078Default scoring metric is accuracy. You can pass in alternate scoring functions, including those available through sklearn:
>>> from sklearn.metrics import precision_score, recall_score
>>> precision = clf.score(X_test, y_test, precision_score)
>>> recall = clf.score(X_test, y_test, recall_score)
>>> print(f'precision: {precision} recall: {recall}')
precision: 0.9914..., recall: 0.9953...To perform predictions, use predict:
>>> ripper_clf.predict(new_data)[:5]
[True, True, False, True, False]Predict class probabilities with predict_proba:
>>> ripper_clf.predict_proba(test)
# Pairs of negative and positive class probabilities
array([[0.01212121, 0.98787879],
[0.01212121, 0.98787879],
[0.77777778, 0.22222222],
[0.2 , 0.8 ],
...We can also ask our model to tell us why it made each positive prediction using give_reasons:
>>> ripper_clf.predict(new_data[:5], give_reasons=True)
([True, True, False, True, True]
[<Rule [physician-fee-freeze=n]>],
[<Rule [physician-fee-freeze=n]>,
<Rule [synfuels-corporation-cutback=y^adoption-of-the-budget-resolution=y^anti-satellite-test-ban=n]>], # This example met multiple sufficient conditions for a positive prediction
[],
[<Rule object: [physician-fee-freeze=n]>],
[])wittgenstein is compatible with sklearn model_selection tools such as cross_val_score and GridSearchCV, as well
as ensemblers like StackingClassifier.
Cross validation:
>>> # First dummify your categorical features and booleanize your class values to make sklearn happy
>>> X_train = pd.get_dummies(X_train, columns=X_train.select_dtypes('object').columns)
>>> y_train = y_train.map(lambda x: 1 if x=='p' else 0)
>>> cross_val_score(ripper_clf, X_train, y_train)Grid search:
>>> from sklearn.model_selection import GridSearchCV
>>> param_grid = {"prune_size": [0.33, 0.5], "k": [1, 2]}
>>> grid = GridSearchCV(estimator=ripper, param_grid=param_grid)
>>> grid.fit(X_train, y_train)Ensemble:
>>> from sklearn.ensemble import StackingClassifier
>>> from sklearn.tree import DecisionTreeClassifier
>>> from sklearn.naive_bayes import GaussianNB
>>> from sklearn.linear_model import LogisticRegression
>>> tree = DecisionTreeClassifier(random_state=42)
>>> nb = GaussianNB(random_state=42)
>>> estimators = [("rip", ripper_clf), ("tree", tree), ("nb", nb)]
>>> ensemble_clf = StackingClassifier(estimators=estimators, final_estimator=LogisticRegression())
>>> ensemble_clf.fit(X_train, y_train)You can directly specify a new model, modify a preexisting model, or train from a preexisting model -- whether to take into account subject matter expertise, to create a baseline for scoring, or for insight into what the model is doing.
To specify a new model, use init_ruleset:
>>> ripper_clf = RIPPER(random_state=42)
>>> ripper_clf.init_ruleset("[[Cap-shape=x^Cap-color=n] V [Odor=c] V ...]", class_feat=..., pos_class=...)
>>> ripper_clf.predict(df)
...To modify a preexisting model, use add_rule, replace_rule, remove_rule, or insert_rule. To alter a model by index, use replace_rule_at, remove_rule_at, or insert_rule_at:
>>> ripper_clf.replace_rule_at(1, '[Habitat=l]')
>>> ripper_clf.insert_rule(insert_before_rule='[Habitat=l]', new_rule='[Gill-size=n ^ Gill-color=b]')
>>> ripper_clf.out_model()
[[delicious=y^spooky-looking=y] V
[Gill-size=n ^ Gill-color=b] V
[Habitat=l]]To specify a starting point for training, use initial_model when calling fit:
>>> ripper_clf.fit(
>>> X_train,
>>> y_train,
>>> initial_model="[[delicious=y^spooky-looking=y] V [Odor=c]]")Expected string syntax for a Ruleset is [<Rule1> V <Rule2> V ...], for a Rule [<Cond1>^<Cond2>^...], and for a Cond feature=value`. '^' represents 'and'; 'V' represents 'or'. (See the Training section above).
Use the interpret module to interpret non-wittgenstein models. interpret_model generates a ruleset that approximates some black-box model. It does to by fitting a wittgenstein classifier to the predictions of the other model.
# Train the model we want to interpret
>>> from tensorflow.keras import Sequential
>>> from tensorflow.keras.layers import Dense
>>> mlp = Sequential()
>>> mlp.add(Dense(60, input_dim=13, activation='relu'))
>>> mlp.add(Dense(30, activation='relu'))
>>> mlp.add(Dense(1, activation='sigmoid'))
>>> mlp.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
>>> mlp.fit(
>>> X_train,
>>> y_train,
>>> batch_size=1,
>>> epochs=10)
# Create and fit wittgenstein classifier to use as a model interpreter.
>>> from wittgenstein.interpret import interpret_model, score_fidelity
>>> interpreter = RIPPER(random_state=42)
>>> interpret_model(model=mlp, X=X_train, interpreter=interpreter).out_pretty()
[[Proline=>1227.0] V
[Proline=880.0-1048.0] V
[Proline=1048.0-1227.0] V
[Proline=736.0-880.0] V
[Alcalinityofash=16.8-17.72]]We can also use the now-fitted interpreter to approximate the reasons behind the underlying model's positive predictions. (See Prediction).
>>> preds = (mlp.predict(X_test.tail()) > .5).flatten()
>>> _, interpretation = interpreter.predict(X_test.tail(), give_reasons=True)
>>> print(f'tf preds: {preds}\n')
>>> interpretation
tf preds: [ True False False True False]
[[<Rule [Proline=880.0-1048.0]>],
[],
[],
[<Rule [Proline=736.0-880.0]>, <Rule [Alcalinityofash=16.8-17.72]>],
[]]Score how faithfully the interpreter fits the underlying model with score_fidelity.
>>> score_fidelity(
>>> X_test,
>>> interpreter,
>>> model=mlp,
>>> score_function=[precision_score, recall_score, f1_score])
[1.0, 0.7916666666666666, 0.8837209302325582]If you encounter any issues, or if you have feedback or improvement requests for how wittgenstein could be more helpful for you, please post them to issues, and I'll respond.
Contributions are welcome! If you are interested in contributing, let me know at ilan.moscovitz@gmail.com or on linkedin.
- My article in Towards Data Science explaining IREP, RIPPER, and wittgenstein
- Furnkrantz-Widmer IREP paper
- Cohen's RIPPER paper
- Partial decision trees
- Bayesian Rulesets
- C4.5 paper including all the gory details on MDL
- Philosophical Investigations
- Improvements to predict_proba calculation, including smoothing
- Speedup for binning continuous features (~several orders of magnitude)
- Add support for expert feedback: Ability to explicitly specify and alter models.
- Add surrogate interpreter
- Add support for non-pandas datasets (ex. numpy arrays)
- Minor bugfixes and optimizations
- Algorithmic optimizations to improve training speed (~10x - ~100x)
- Support for training on iterable datatypes besides DataFrames, such as numpy arrays and python lists
- Compatibility with sklearn ensembling metalearners and sklearn model_selection
.predict_probareturns probas in neg, pos order- Certain parameters (hyperparameters, random_state, etc.) should now be passed into IREP/RIPPER constructors rather than the .fit method.
- Sundry bugfixes