Pandas dataset into an array for modelling in Scikit-Learn

Question

Can we run scikit-learn models on Pandas DataFrames or do we need to convert DataFrames into NumPy arrays?

Answer 1

You can use pandas.DataFrame with sklearn, for example:

import pandas as pd
from sklearn.cluster import KMeans

data = [(0.2, 10),
        (0.3, 12),
        (0.24, 14),
        (0.8, 30),
        (0.9, 32),
        (0.85, 33.3),
        (0.91, 31),
        (0.1, 15),
        (-0.23, 45)]

p_df = pd.DataFrame(data)
kmeans = KMeans(init='k-means++', n_clusters=3, n_init=10)
kmeans.fit(p_df)

Result:

>>> kmeans.labels_
array([0, 0, 0, 2, 2, 2, 2, 0, 1], dtype=int32)

Answer 2

Pandas DataFrames are very good at acting like Numpy arrays when they need to. If in doubt, you can always use the values attribute to get a Numpy representation (df.values will give you a Numpy array of the values in DataFrame df.