scikit-learn/examples/linear_model/plot_ransac.py

67 lines
2.0 KiB
Python

"""
===========================================
Robust linear model estimation using RANSAC
===========================================
In this example we see how to robustly fit a linear model to faulty data using
the RANSAC algorithm.
"""
import numpy as np
from matplotlib import pyplot as plt
from sklearn import linear_model, ensemble
# Set random seed for both equal data noise and equal random sample selection
np.random.seed(seed=1)
# Generate coordinates of line
X = np.arange(-200, 200)
y = 0.2 * X + 20
data = np.column_stack([X, y])
# Add faulty data
faulty = np.array(30 * [(180, -100)], dtype=np.double)
faulty += 5 * np.random.normal(size=faulty.shape)
data[:faulty.shape[0]] = faulty
# Add gaussian noise to coordinates
noise = np.random.normal(size=data.shape)
data += 0.5 * noise
data[::2] += 5 * noise[::2]
data[::4] += 20 * noise[::4]
X = data[:, 0][:, np.newaxis]
y = data[:, 1]
# Fit line using all data
model = linear_model.LinearRegression()
model.fit(X, y)
# Robustly fit linear model with RANSAC algorithm
model_ransac = linear_model.RANSAC(linear_model.LinearRegression())
model_ransac.fit(X, y)
inlier_mask = model_ransac.inlier_mask_
outlier_mask = np.logical_not(inlier_mask)
# Robustly fit linear model with bagged linear regressor
model_bagged = ensemble.BaggingRegressor(linear_model.LinearRegression())
model_bagged.fit(X, y)
# Generate coordinates of estimated models
line_X = np.arange(-250, 250)
line_y = model.predict(line_X[:, np.newaxis])
line_y_ransac = model_ransac.predict(line_X[:, np.newaxis])
line_y_bagged = model_bagged.predict(line_X[:, np.newaxis])
plt.plot(data[inlier_mask, 0], data[inlier_mask, 1], '.g',
label='RANSAC inliers')
plt.plot(data[outlier_mask, 0], data[outlier_mask, 1], '.r',
label='RANSAC outliers')
plt.plot(line_X, line_y, '-k', label='Linear regressor')
plt.plot(line_X, line_y_ransac, '-b', label='RANSAC regressor')
plt.plot(line_X, line_y_bagged, '-y', label='Bagging regressor')
plt.legend(loc='lower left')
plt.show()