"""
==========================================================================
Illustration of the sample selection for the different NearMiss algorithms
==========================================================================

This example illustrates the different way of selecting example in NearMiss.

"""

import matplotlib.pyplot as plt
import numpy as np

from sklearn.neighbors import NearestNeighbors

print(__doc__)

rng = np.random.RandomState(18)

###############################################################################
# This function allows to make nice plotting


def make_plot_despine(ax):
    ax.spines['top'].set_visible(False)
    ax.spines['right'].set_visible(False)
    ax.get_xaxis().tick_bottom()
    ax.get_yaxis().tick_left()
    ax.spines['left'].set_position(('outward', 10))
    ax.spines['bottom'].set_position(('outward', 10))
    ax.set_xlim([0.5, 3.])
    ax.set_ylim([0.5, 3.])
    ax.set_xlabel(r'$X_1$')
    ax.set_ylabel(r'$X_2$')
    ax.legend()


###############################################################################
# We can start by generating some data to later illustrate the principle of
# each NearMiss heuritic rules.

# minority class
X_minority = np.transpose([[1.1, 1.3, 1.15, 0.8, 0.8, 0.6, 0.55],
                           [1., 1.5, 1.7, 2.5, 2.0, 1.2, 0.55]])
# majority class
X_majority = np.transpose([[2.1, 2.12, 2.13, 2.14, 2.2, 2.3, 2.5, 2.45],
                           [1.5, 2.1, 2.7, 0.9, 1.0, 1.4, 2.4, 2.9]])

plt.ioff()
fig, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(20, 6))

# plot the majority and minority samples
for ax in (ax1, ax2, ax3):
    ax.scatter(X_minority[:, 0], X_minority[:, 1],
               label='Minority class', s=200, marker='_')
    ax.scatter(X_majority[:, 0], X_majority[:, 1],
               label='Majority class', s=200, marker='+')

###############################################################################
# NearMiss-1
###############################################################################

###############################################################################
# NearMiss-1 selects samples from the majority class for which the average
# distance to some nearest neighbours is the smallest. In the following
# example, we use a 3-NN to compute the average distance on 2 specific samples
# of the majority class. Therefore, in this case the point linked by the
# green-dashed line will be selected since the average distance is smaller.

nearest_neighbors = NearestNeighbors(n_neighbors=3)
nearest_neighbors.fit(X_minority)
dist, ind = nearest_neighbors.kneighbors(X_majority[:2, :])
# Compute the average distance
dist_avg = dist.sum(axis=1) / 3

for positive_idx, (neighbors, distance, color) in enumerate(
        zip(ind, dist_avg, ['g', 'r'])):
    for make_plot, sample_idx in enumerate(neighbors):
        ax1.plot([X_majority[positive_idx, 0], X_minority[sample_idx, 0]],
                 [X_majority[positive_idx, 1], X_minority[sample_idx, 1]],
                 '--' + color, alpha=0.3,
                 label='Avg. dist.={0:.2f}'.format(distance)
                 if make_plot == 0 else "")
ax1.set_title('NearMiss-1')
make_plot_despine(ax1)
plt.show()