Isolation Forest : Categorical data

Question

I am trying to detect anomalies in a breast cancer dataset using Isolation Forest in sklearn. I am trying to apply Iolation Forest to a mixed data set and it gives me value errors when I fit the model.

This is my dataset : https://archive.ics.uci.edu/ml/machine-learning-databases/breast-cancer/

This is my code :

from sklearn.model_selection import train_test_split
rng = np.random.RandomState(42)

X = data_cancer.drop(['Class'],axis=1)
y = data_cancer['Class'] 

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 20)
X_outliers = rng.uniform(low=-4, high=4, size=(X.shape[0], X.shape[1]))

clf = IsolationForest()
clf.fit(X_train)

This is the error I get :

ValueError: could not convert string to float: '30-39'

Is it possible to use Isolation Forest on categorical data? If yes, how do I do so?

Answer 1

You should encode your categorical data to numerical representation.

There are many ways to encode categorical data, but I suggest that you start with

sklearn.preprocessing.LabelEncoder if cardinality is high and sklearn.preprocessing.OneHotEncoder if cardinality is low.

Here a usage example:

import numpy as np
from numpy import argmax
from sklearn.preprocessing import LabelEncoder
from sklearn.preprocessing import OneHotEncoder
# define example
data = ['cold', 'cold', 'warm', 'cold', 'hot', 'hot', 'warm', 'cold', 'warm', 'hot']
values = np.array(data)
print(values)
# integer encode
label_encoder = LabelEncoder()
integer_encoded = label_encoder.fit_transform(values)
print(integer_encoded)
# binary encode
onehot_encoder = OneHotEncoder(sparse=False)
integer_encoded = integer_encoded.reshape(len(integer_encoded), 1)
onehot_encoded = onehot_encoder.fit_transform(integer_encoded)
print(onehot_encoded)
# invert first example
inverted = label_encoder.inverse_transform([argmax(onehot_encoded[0, :])])
print(inverted)

Output:

['cold' 'cold' 'warm' 'cold' 'hot' 'hot' 'warm' 'cold' 'warm' 'hot']
 
[0 0 2 0 1 1 2 0 2 1]
 
[[ 1.  0.  0.]
 [ 1.  0.  0.]
 [ 0.  0.  1.]
 [ 1.  0.  0.]
 [ 0.  1.  0.]
 [ 0.  1.  0.]
 [ 0.  0.  1.]
 [ 1.  0.  0.]
 [ 0.  0.  1.]
 [ 0.  1.  0.]]
 
['cold']