What is clf and legend in ‘plot_decision_regions’?

from mlxtend.plotting import plot_decision_regions

def knn_comparision(data, k):
    X = data[['x1','x2']].values
    y = data['y'].astype(int).values
    clf = neighbors.KNeighborsClassifier(n_neighbors=k)
    clf.fit(X, y)

    # Plotting decision regions
    plot_decision_regions(X, y, clf=clf, legend=2)

JavaScript
​x
 
from mlxtend.plotting import plot_decision_regions
​
def knn_comparision(data, k):
    X = data[['x1','x2']].values
    y = data['y'].astype(int).values
    clf = neighbors.KNeighborsClassifier(n_neighbors=k)
    clf.fit(X, y)
​
    # Plotting decision regions
    plot_decision_regions(X, y, clf=clf, legend=2)
​

What are the parameters ‘clf‘ and ‘legend‘ in plot_decision_regions?

Answer

clf is the classifier object being returned from neighbors.KNeighborsClassifier, which is likely coming from sklearn.

BigBen linked the documentation already for the plot_decision_regions function, which explains what they do.

Advertisement

Answer