def create_plot(tree, name="tree"):
fig = plt.figure(1, facecolor="white")
fig.clf()
create_plot.ax1 = plt.subplot(111, frameon=False, xticks=[], yticks=[])
plot_tree.total_w = float(get_leafs_number(tree))
plot_tree.total_d = float(get_tree_depth(tree))
plot_tree.x_off = -0.5 / plot_tree.total_w
plot_tree.y_off = 1.0
plot_tree(tree, (0.5, 1.0), "")
fig.savefig(data_out_path("decision_tree/{}.png").format(name))
def test_dating_set():
ho_ratio = 0.1
filepath = data_path('datingTestSet2.txt')
data_set, labels = import_dataset(filepath)
norm_data_set, ranges, min_vals = normalize(data_set)
m = norm_data_set.shape[0]
num_test = int(m * ho_ratio)
error_count = 0.0
for i in range(num_test):
classifier_result = classify(
norm_data_set[i, :],
norm_data_set[num_test:m, :],
labels[num_test:m],
7
)
if classifier_result != labels[i]:
error_count += 1.0
print 'the total error rate is: %f' % (error_count / num_test)
colors = []
sizes = 15.0 * array(labels)
for label in labels:
if label == 1:
color = 'red'
elif label == 2:
color = 'green'
elif label == 3:
color = 'blue'
else:
color = 'gray'
colors.append(color)
_fig = plt.figure()
_ax = _fig.add_subplot(111)
type1 = _ax.scatter([-10], [-10], s=15, c='red')
type2 = _ax.scatter([-10], [-15], s=30, c='green')
type3 = _ax.scatter([-10], [-20], s=45, c='blue')
fig = plt.figure()
ax = fig.add_subplot(111)
ax.scatter(data_set[:, 0], data_set[:, 1], c=colors, s=sizes)
ax.set_xlabel('Frequent Flyier Miles Earned Per Year')
ax.set_ylabel('Percentage of Time Spent Playing Video Games')
ax.legend(
[type1, type2, type3],
['Did Not Like', 'Liked in Small Doses', 'Liked in Large Doses'],
loc=2
)
ax.grid(True)
fig.savefig(data_out_path('games_vs_miles.png'))
fig = plt.figure()
ax = fig.add_subplot(111)
ax.scatter(data_set[:, 1], data_set[:, 2], c=colors, s=sizes)
ax.set_xlabel('Percentage of Time Spent Playing Video Games')
ax.set_ylabel('Liters of Ice Cream Consumed Per Week')
ax.legend(
[type1, type2, type3],
['Did Not Like', 'Liked in Small Doses', 'Liked in Large Doses'],
loc=4
)
ax.grid(True)
fig.savefig(data_out_path('ice_cream_vs_games.png'))
def load_tree(filename):
with open(data_out_path(filename), 'r') as f:
return pickle.load(f)
def dump_tree(tree, filename):
with open(data_out_path(filename), 'w') as f:
pickle.dump(tree, f)