Hadrien Jean hadrienj

## utils_essential_math_data_science.py
import matplotlib.pyplot as plt
import numpy as np
import matplotlib.collections as mcoll


def matrix_2d_effect(transfo_matrix, vectorsCol=["#FFD800", "#00CD79"]):
    """
    Modify the unit circle and basis vector by applying a matrix.
    Visualize the effect of the matrix in 2D.

## unit_vectors.py
# choose figure size
plt.figure(figsize=(6, 6))

# Assure that ticks are displayed with a step equal to 1
ax = plt.gca()
ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
ax.yaxis.set_major_locator(ticker.MultipleLocator(1))

# draw axes
plt.axhline(0, c='#A9A9A9', zorder=0)

## preprocessing-data-for-machine-learning-and-deep-learning_code35.py
plotImage(X[12, :])
plotImage(X_ZCA_rescaled[12, :])

## preprocessing-data-for-machine-learning-and-deep-learning_code34.py
X_ZCA_rescaled = (X_ZCA - X_ZCA.min()) / (X_ZCA.max() - X_ZCA.min())
print 'min:', X_ZCA_rescaled.min()
print 'max:', X_ZCA_rescaled.max()

## preprocessing-data-for-machine-learning-and-deep-learning_code33.py
epsilon = 0.1
X_ZCA = U.dot(np.diag(1.0/np.sqrt(S + epsilon))).dot(U.T).dot(X_norm)

plotImage(X[12, :])
plotImage(X_ZCA[12, :])

## preprocessing-data-for-machine-learning-and-deep-learning_code33.py
print np.diag(S)
print '\nshape:', np.diag(S).shape

## preprocessing-data-for-machine-learning-and-deep-learning_code32.py
print U.shape, S.shape

## preprocessing-data-for-machine-learning-and-deep-learning_code31.py
U,S,V = np.linalg.svd(cov)

## preprocessing-data-for-machine-learning-and-deep-learning_code30.py
cov = np.cov(X_norm, rowvar=True)

## preprocessing-data-for-machine-learning-and-deep-learning_code29.py
X_norm.mean(axis=0)
	import matplotlib.pyplot as plt
	import numpy as np
	import matplotlib.collections as mcoll


	def matrix_2d_effect(transfo_matrix, vectorsCol=["#FFD800", "#00CD79"]):
	"""
	Modify the unit circle and basis vector by applying a matrix.
	Visualize the effect of the matrix in 2D.
	# choose figure size
	plt.figure(figsize=(6, 6))

	# Assure that ticks are displayed with a step equal to 1
	ax = plt.gca()
	ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
	ax.yaxis.set_major_locator(ticker.MultipleLocator(1))

	# draw axes
	plt.axhline(0, c='#A9A9A9', zorder=0)
	X_ZCA_rescaled = (X_ZCA - X_ZCA.min()) / (X_ZCA.max() - X_ZCA.min())
	print 'min:', X_ZCA_rescaled.min()
	print 'max:', X_ZCA_rescaled.max()
	epsilon = 0.1
	X_ZCA = U.dot(np.diag(1.0/np.sqrt(S + epsilon))).dot(U.T).dot(X_norm)

	plotImage(X[12, :])
	plotImage(X_ZCA[12, :])