Handwritten digits and Locally Linear Embedding

plot_lle_digits.py

"""
===============================================
Handwritten digits and Locally Linear Embedding 
===============================================

An illustration of locally linear embedding on the digits dataset.
"""

# Author: Fabian Pedregosa -- <fabian.pedregosa@inria.fr>
# License: BSD, (C) INRIA 2011

print __doc__

import numpy as np
import pylab as pl
from matplotlib.offsetbox import AnnotationBbox, OffsetImage

#----------------------------------------------------------------------
# Locally linear embedding of the digits dataset

from scikits.learn import manifold, metrics, datasets
digits = datasets.load_digits(n_class=6)

print "Computing LLE embedding"
X_r, err = manifold.locally_linear_embedding(digits.data, 30, 2, reg=1e-2)
print "Done. Reconstruction error: ", err


#----------------------------------------------------------------------
# Scale and visualize the embedding vectors

x_min, x_max = np.min(X_r, 0), np.max(X_r, 0)
X_r = (X_r - x_min) / (x_max - x_min)

ax = pl.subplot(111)
for i in range(digits.data.shape[0]):
    pl.text(X_r[i, 0], X_r[i, 1], str(digits.target[i]),
            color=pl.cm.Set1(digits.target[i] / 10.),
            fontdict={'weight': 'bold', 'size' : 9})

shown_images = np.array([[1., 1.]]) # just something big
for i in range(digits.data.shape[0]):
    dist = np.sum((X_r[i] - shown_images)**2, 1)
    if np.min(dist) < 4e-3:
        # don't show points that are too close
        continue
    shown_images = np.r_[shown_images, [X_r[i]]]
    imagebox = AnnotationBbox(
        OffsetImage(digits.images[i], cmap=pl.cm.gray_r),
        X_r[i])
    ax.add_artist(imagebox)

pl.xticks([]), pl.yticks([])
pl.show()