def plot_components_2(X_projected, Y_projected, which_dimensions_to_plot, images=None, ax=None, image_scale=1, markersize=10, thumb_frac=0.05, cmap='gray'):
# https://jakevdp.github.io/PythonDataScienceHandbook/05.10-manifold-learning.html
# X_projected: rows are features and columns are samples
# which_dimensions_to_plot: a list of two integers, index starting from 0
X_projected = np.vstack((X_projected[which_dimensions_to_plot[0], :], X_projected[which_dimensions_to_plot[1], :])) #--> only take two dimensions to plot
X_projected = X_projected.T
ax = ax or plt.gca()
Y_projected = Y_projected.ravel()
ax.plot(X_projected[Y_projected == 0, 0], X_projected[Y_projected == 0, 1], '.k', markersize=markersize)
ax.plot(X_projected[Y_projected == 1, 0], X_projected[Y_projected == 1, 1], '.r', markersize=markersize)
images = resize(images, (images.shape[0], images.shape[1]*image_scale, images.shape[2]*image_scale), order=5, preserve_range=True, mode="constant")
if images is not None:
min_dist_2 = (thumb_frac * max(X_projected.max(0) - X_projected.min(0))) ** 2
shown_images = np.array([2 * X_projected.max(0)])
for i in range(X_projected.shape[0]):
dist = np.sum((X_projected[i] - shown_images) ** 2, 1)
if np.min(dist) < min_dist_2:
# don't show points that are too close
continue
shown_images = np.vstack([shown_images, X_projected[i]])
if Y_projected[i] == 0:
imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(images[i], cmap=cmap), X_projected[i])
else:
imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(images[i], cmap=cmap), X_projected[i], bboxprops =dict(edgecolor='red', lw=3))
ax.add_artist(imagebox)
# # plot the first (original) image once more to be on top of other images:
# # change color of frame (I googled: python OffsetImage highlight frame): https://stackoverflow.com/questions/40342379/show-images-in-a-plot-using-matplotlib-with-a-coloured-frame
# imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(images[0], cmap=cmap), X_projected[0], bboxprops =dict(edgecolor='red'))
# ax.add_artist(imagebox)
plt.xlabel("dimension " + str(which_dimensions_to_plot[0] + 1), fontsize=13)
plt.ylabel("dimension " + str(which_dimensions_to_plot[1] + 1), fontsize=13)
plt.show()
def standardizePlot(ax,
title,
xlab,
ylab,
xlim=None,
ylim=None,
hline=True,
legend=True,
lblpressure=True,
putimg=True,
compliance=False):
ax.axhline(linewidth=1, color='k', zorder=1)
ax.set_xlabel(xlab)
ax.set_ylabel(ylab)
ax.set_title(title)
if xlim is not None: ax.set_xlim(xlim)
if ylim is not None: ax.set_ylim(ylim)
if legend: ax.legend(prop={'size': 8})
if lblpressure:
ax.text(.15,
-.2,
"(ME Pressure)",
ha='center',
transform=axs.transAxes)
ax.text(.85,
-.2,
"(IE Pressure)",
ha='center',
transform=axs.transAxes)
if putimg:
if compliance:
MEy = 0.26
IEy = 0.26
else:
MEy = 0.56
IEy = 0.44
arr_img = plt.imread(pjoin(Base_Folder, "MembraneForPressureME.png"),
format='png')
imagebox = mplob.OffsetImage(arr_img, zoom=0.08)
imagebox.image.axes = ax
ab = mplob.AnnotationBbox(imagebox, (0.15, MEy),
xycoords='axes fraction',
pad=0.5,
frameon=False)
ax.add_artist(ab)
arr_img = plt.imread(pjoin(Base_Folder, "MembraneForPressureIE.png"),
format='png')
imagebox = mplob.OffsetImage(arr_img, zoom=0.08)
imagebox.image.axes = ax
ab = mplob.AnnotationBbox(imagebox, (0.85, IEy),
xycoords='axes fraction',
pad=0.5,
frameon=False)
ax.add_artist(ab)
def plot_embedding(X,
imagepaths,
collection_size,
title=None,
test_image=False,
image_size=42):
'''
Plots the images in the embedding space. The test image has red border.
X: ndarray of float
The distance matrix (2x2 for the 2D visualization).
imagepaths: list of str
A list containing all paths to the images.
collection_size: int
The number of images inside the collection.
test_image: bool, default False
If True, then we assume that the last vector in X corresponds to a
test image, so we annotate it with a red box and the others with a
black one. Otherwise, we annotate all images with a black box.
image_size: int, default 42
The final size of the thumbnails to be shown in the plot will be
(image_size, image_size).
Kudos to georgeretsi for the code.
'''
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
#for i in range(X.shape[0]):
# plt.text(X[i, 0], X[i, 1], str(names[i]),fontdict={'weight': 'bold', 'size': 9})
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
#dist = np.sum((X[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[i]]]
timg = cv2.imread(imagepaths[i])
timg = timg[:, :, ::-1] # BGR -> RGB
timg = cv2.resize(timg, dsize=(image_size, image_size))
if (i == X.shape[0] - 1) and (test_image == True):
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(timg, cmap=plt.cm.gray_r),
X[i],
bboxprops=dict(edgecolor='red'))
else:
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(timg, cmap=plt.cm.gray_r),
X[i],
bboxprops=dict(edgecolor='black'))
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
plt.show(block=True)
def plot_embedding_annotation(emb, digits, threshold, ax=None, rescale=True):
"""
This function is used to visualize the learned low-d features
We intend to see if we learn to disentangle factors of variations
@emb : the input low-d feature
@digits : the immage annotation of emb
@threshold: minimal distances between two points
"""
# Rescaling
if rescale:
x_min, x_max = np.min(emb, 0), np.max(emb, 0)
emb = (emb - x_min) / (x_max - x_min)
_, ax = plt.subplots()
if hasattr(offsetbox, 'AnnotationBbox'):
mycanvas = np.array([[1., 1.]])
for i in range(digits.shape[0]):
dist = np.sum((emb[i] - mycanvas)**2, 1)
if np.min(dist) < threshold:
# don't show points that are too close
# You may try different threshold
continue
mycanvas = np.r_[mycanvas, [emb[i]]]
imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(
digits[i], cmap=plt.cm.gray_r),
emb[i],
frameon=False)
ax.add_artist(imagebox)
ax.set_xticks([])
ax.set_yticks([])
return 0
def visualize(embed, x_test, y_test):
feat = embed
ax_min = np.min(embed, 0)
ax_max = np.max(embed, 0)
ax_dist_sq = np.sum((ax_max - ax_min)**2)
plt.figure()
ax = plt.gca()
ax.grid(False)
ax = plt.subplot(111)
colormap = plt.get_cmap('tab10')
shown_images = np.array([[1., 1.]])
for i in range(feat.shape[0]):
dist = np.sum((feat[i] - shown_images)**2, 1)
if np.min(
dist
) < 3e-4 * ax_dist_sq: # don't show points that are too close
continue
shown_images = np.r_[shown_images, [feat[i]]]
patch_to_color = np.expand_dims(x_test[i], -1)
patch_to_color = np.tile(patch_to_color, (1, 1, 3))
patch_to_color = (1 - patch_to_color) * (
1, 1, 1) + patch_to_color * colormap(y_test[i] / 10.)[:3]
imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(
patch_to_color, zoom=0.5, cmap=plt.cm.gray_r),
xy=feat[i],
frameon=False)
ax.add_artist(imagebox)
plt.axis([ax_min[0], ax_max[0], ax_min[1], ax_max[1]])
plt.title('Embedding from the last layer of the network')
plt.show()
def visualize(embed, x_test):
# two ways of visualization: scale to fit [0,1] scale
# feat = embed - np.min(embed, 0)
# feat /= np.max(feat, 0)
# two ways of visualization: leave with original scale
feat = embed
print('embed:',embed)
ax_min = np.min(embed,0)
ax_max = np.max(embed,0)
print("max and min:",ax_max,ax_min )
ax_dist_sq = np.sum((ax_max-ax_min)**2)
plt.figure()
ax = plt.subplot(111)
shown_images = np.array([[1., 1.]])
for i in range(feat.shape[0]):
dist = np.sum((feat[i] - shown_images)**2, 1)
if np.min(dist) < 3e-4*ax_dist_sq: # don't show points that are too close
continue
shown_images = np.r_[shown_images, [feat[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(x_test[i], zoom=0.6, cmap=plt.cm.gray_r),
xy=feat[i], frameon=False
)
ax.add_artist(imagebox)
plt.axis([ax_min[0], ax_max[0], ax_min[1], ax_max[1]])
# plt.xticks([]), plt.yticks([])
plt.title('Embedding from the last layer of the network')
plt.show()
def acquire_playerPic(player_id, zoom, offset=(250, 370)):
try:
img_path = os.getcwd() + '/' + str(player_id) + '.png'
player_pic = plt.imread(img_path)
except (ValueError, IOError):
try:
pic = urllib.urlretrieve(
"http://stats.nba.com/media/players/230x185/" +
str(player_id) + ".png",
str(player_id) + ".png")
player_pic = plt.imread(pic[0])
except (ValueError, IOError):
try:
pic = urllib.urlretrieve(
"https://ak-static.cms.nba.com/wp-content/uploads/headshots/nba/latest/260x190/"
+ str(player_id) + ".png",
str(player_id) + ".png")
player_pic = plt.imread(pic[0])
except (ValueError, IOError):
img_path = os.getcwd() + '/chart_icon.png'
player_pic = plt.imread(img_path)
img = osb.OffsetImage(player_pic, zoom)
img = osb.AnnotationBbox(img,
offset,
xycoords='data',
pad=0.0,
box_alignment=(1, 0),
frameon=False)
return img
def plot_embedding(Xp, y, imgs, title=None, figsize=(12, 4)):
x_min, x_max = numpy.min(Xp, 0), numpy.max(Xp, 0)
X = (Xp - x_min) / (x_max - x_min)
fig, ax = plt.subplots(1, 2, figsize=figsize)
for i in range(X.shape[0]):
ax[0].text(X[i, 0],
X[i, 1],
str(y[i]),
color=plt.cm.Set1(y[i] / 10.),
fontdict={
'weight': 'bold',
'size': 9
})
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = numpy.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
dist = numpy.sum((X[i] - shown_images)**2, 1)
if numpy.min(dist) < 4e-3:
# don't show points that are too close
continue
shown_images = numpy.r_[shown_images, [X[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(imgs[i], cmap=plt.cm.gray_r), X[i])
ax[0].add_artist(imagebox)
ax[0].set_xticks([]), ax[0].set_yticks([])
ax[1].plot(Xp[:, 0], Xp[:, 1], '.')
if title is not None:
ax[0].set_title(title)
return ax
def plot_mnist(X, y, X_embedded, name, min_dist=10.0):
fig = figure(figsize=(10, 10))
ax = axes(frameon=False)
title("\\textbf{MNIST dataset} -- Two-dimensional "
"embedding of 70,000 handwritten digits with %s" % name)
setp(ax, xticks=(), yticks=())
subplots_adjust(left=0.0,
bottom=0.0,
right=1.0,
top=0.9,
wspace=0.0,
hspace=0.0)
scatter(X_embedded[:, 0], X_embedded[:, 1], c=y, marker="x")
if min_dist is not None:
from matplotlib import offsetbox
shown_images = array([[15., 15.]])
indices = arange(X_embedded.shape[0])
random.shuffle(indices)
for i in indices[:5000]:
dist = sum((X_embedded[i] - shown_images)**2, 1)
if min(dist) < min_dist:
continue
shown_images = np.r_[shown_images, [X_embedded[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(X[i].reshape(28, 28), cmap=cm.gray_r),
X_embedded[i])
ax.add_artist(imagebox)
def plot_tsne_3D(X_tsne, images, labels, azim=120, distance=70000):
fig = plt.figure(figsize=(20, 20))
ax = fig.add_subplot(111, projection=Axes3D.name)
ax2d = fig.add_subplot(111, frame_on=False)
ax2d.axis("off")
ax.view_init(elev=30., azim=azim)
for i in range(X_tsne.shape[0]):
ax.scatter(X_tsne[i, 0],
X_tsne[i, 1],
X_tsne[i, 2],
c=plt.cm.Set1(labels[i] / 10.),
s=100)
if hasattr(offsetbox, 'AnnotationBbox'):
shown_images = np.array([[1., 1., 1.]])
for i in range(images.shape[0]):
dist = np.sum((X_tsne[i] - shown_images)**2, 1)
if np.min(dist) < distance:
# don't show points that are too close
continue
shown_images = np.r_[shown_images, [X_tsne[i]]]
image = Image.fromarray(images[i].reshape(28, 28), 'L')
#inverted_image = PIL.ImageOps.invert(image)
image.thumbnail((40, 40), Image.ANTIALIAS)
imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(image),
proj(X_tsne[i], ax, ax2d))
ax2d.add_artist(imagebox)
ax.set_xlabel('X Label')
ax.set_ylabel('Y Label')
ax.set_zlabel('Z Label')
plt.title('t-SNE over the Fashion-MNIST')
plt.savefig("/tmp/fashion/movie%d.png" % azim)
def plot_embedding(X,y,images, title=None):
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
ly = float(len(set(y)))
for i in range(X.shape[0]):
circle = plt.Circle((X[i,0],X[i,1]),0.001,color=plt.cm.Set1(y[i]/ly))
ax.add_artist(circle)
# plt.text(X[i, 0], X[i, 1], str(y[i]),
# color=plt.cm.Set1(y[i] / 10.),
# fontdict={'weight': 'bold', 'size': 9})
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
dist = np.sum((X[i] - shown_images) ** 2, 1)
if np.min(dist) < 4e-4:
# don't show points that are too close
continue
shown_images = np.r_[shown_images, [X[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(st.resize(plt.imread(os.path.join(args.image_root,images[i])),(64,32)), cmap=plt.cm.gray_r),
X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
def plot_components(data,
model,
images=None,
ax=None,
thumb_frac=0.05,
cmap='gray'):
ax = ax or plt.gca()
proj = model.fit_transform(data)
ax.plot(proj[:, 0], proj[:, 1], '.k')
if images is not None:
min_dist_2 = (thumb_frac * max(proj.max(0) - proj.min(0)))**2
shown_images = np.array([2 * proj.max(0)])
for i in range(0, data.shape[0]):
dist = np.sum((proj[i] - shown_images)**2, 1)
if np.min(dist) < min_dist_2:
# dont show points that are too close
continue
shown_images = np.vstack([shown_images, proj[i]])
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(images[i], cmap=cmap), proj[i])
ax.add_artist(imagebox)
def plotScatter(X, title,minD):
cm = plt.cm.get_cmap('RdYlBu')
plt.scatter(X[:, 0], X[:, 1], c=trainLabel, s=3,cmap=cm)
plt.colorbar()
plt.title(title)
plt.show()
plt.figure()
ax = plt.subplot(111)
cm = plt.cm.get_cmap('RdYlBu')
plt.scatter(X[:, 0], X[:, 1], c=trainLabel, s=3,cmap=cm)
plt.colorbar()
shown_images = np.array([[1., 1.]])
for i in range(trainImages.shape[0]):
dist = np.sum((X[i] - shown_images) ** 2, 1)
if np.min(dist) < minD:
continue
shown_images = np.r_[shown_images, [X[i]]]
if(trainLabel[i]==2):
imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(trainImages[i].reshape(28, 28)),X[i])
ax.add_artist(imagebox)
plt.title(title)
plt.show()
def plot_embedding(X, labs, title=None):
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.text(X[i, 0],
X[i, 1], ('x' if labs[i] == 0 else 'o'),
color=('green' if labs[i] == 0 else 'red'),
fontdict={
'weight': 'bold',
'size': 9
})
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(digits.data.shape[0]):
dist = np.sum((X[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[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(digits.images[i], cmap=plt.cm.gray_r),
X[i],
bboxprops={'edgecolor': ('green' if labs[i] == 0 else 'red')})
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
def plot_embedding_v2(X, X_origin, title=None, dims=[None, 28, 28]):
dims[0] = X.shape[0]
X_origin = X_origin.values.astype(np.float).reshape(dims)
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.text(X[i, 0], X[i, 1], str(y_train.values[i]),
color=plt.cm.Set1(y_train.values.astype(np.int)[i] / 10.),
fontdict={'weight': 'bold', 'size': 9})
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
dist = np.sum((X[i] - shown_images) ** 2, 1)
if np.min(dist) < 3e-3:
# don't show points that are too close
continue
shown_images = np.r_[shown_images, [X[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(X_origin[i], cmap=plt.cm.gray_r),
X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
plt.show()
def plot_embedding(X,
y,
images,
title=None,
figsize=(20, 20),
img_size=(65, 65)):
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure(figsize=figsize)
ax = plt.subplot(111)
colormap = plt.cm.gist_ncar
colors = [colormap(i) for i in np.linspace(0, 1, len(set(y)))]
for i in range(X.shape[0]):
plt.scatter(X[i, 0], X[i, 1], color=colors[y[i]])
if hasattr(offsetbox, 'AnnotationBbox'):
shown_images = np.array([[1., 1.]]) # just something big
for i in range(images.shape[0]):
dist = np.sum((X[i] - shown_images)**2, 1)
if np.min(dist) < 6e-3:
## don't show points that are too close
continue
shown_images = np.r_[shown_images, [X[i]]]
img = resize(images[i], img_size)
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(img, cmap=plt.cm.gray_r), X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
def plot_embedding(X, label, origin_img, show_origin_image=False):
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.text(X[i, 0],
X[i, 1],
str(label[i]),
color=plt.cm.Set1(label[i] / 10.),
fontdict={
'weight': 'bold',
'size': 9
})
if show_origin_image and hasattr(offsetbox, 'AnnotationBbox'):
shown_images = np.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
dist = np.sum((X[i] - shown_images)**2, 1)
if np.min(dist) < 4e-3:
continue
shown_images = np.r_[shown_images, [X[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(origin_img[i], cmap=plt.cm.gray_r), X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
def plot_components(data,
proj,
images=None,
ax=None,
thumb_frac=0.05,
cmap='copper'):
# scaler = MinMaxScaler(feature_range=(0,255))
ax = ax or plt.gca()
ax.plot(proj[:, 0], proj[:, 1], '.k')
if images is not None:
min_dist_2 = (thumb_frac * max(proj.max(0) - proj.min(0)))**2
shown_images = np.array([2 * proj.max(0)])
for i in range(data.shape[0]):
dist = np.sum((proj[i] - shown_images)**2, 1)
if np.min(dist) < min_dist_2:
# don't show points that are too close
continue
shown_images = np.vstack([shown_images, proj[i]])
# #Rescale images to be 0 to 255
# for channel in range(0,images.shape[1]):
# scaler.fit(images[i,channel])
# scaler.fit_transform(images[i,channel])
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(images[i], zoom=.2, cmap=cmap), proj[i])
ax.add_artist(imagebox)
def plot_digits(X, digits, title=None, plot_box=True):
colorlist = get_colors(10)
# Scale and visualize the embedding vectors
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.text(X[i, 0], X[i, 1], str(digits.target[i]),
color=colorlist[digits.target[i]],
fontdict={'weight': 'medium', 'size': 'smaller'})
if plot_box and hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(digits.data.shape[0]):
dist = np.sum((X[i] - shown_images) ** 2, 1)
if np.min(dist) < 4e-2:
# don't show points that are too close
continue
shown_images = np.r_[shown_images, [X[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(digits.images[i], cmap=plt.cm.gray_r),
X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
plt.xlim(-0.05, 1.05)
plt.ylim(-0.05, 1.05)
if title is not None:
plt.title(title)
def plot_embedding(X, y, digits, title=None, annotation=True):
"""
:param X: [b, features]
:param title:
:return:
"""
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
# for each point
for i in range(X.shape[0]):
plt.text(X[i, 0], X[i, 1], str(y[i]),
color=plt.cm.Set1(y[i] / 10.),
fontdict={'weight': 'normal', 'size': 10})
if hasattr(offsetbox, 'AnnotationBbox') and annotation:
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
dist = np.sum((X[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[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(digits.images[i], cmap=plt.cm.gray_r),
X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
def plot_embedding(X, title=None):
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.plot(X[i, 0], X[i, 1])
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(digits.data.shape[0]):
dist = np.sum((X[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[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(digits.images[i], cmap=plt.cm.gray_r),
X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
def plot_embedding(X, title=None):
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.text(X[i, 0],
X[i, 1],
str(Y[i]),
color=plt.cm.Set1(Y[i] / 10.),
fontdict={
'weight': 'bold',
'size': 9
})
if hasattr(offsetbox, 'AnnotationBbox'):
shown_images = np.array([[1., 1.]])
for i in range(X.shape[0]):
dist = np.sum((X[i] - shown_images)**2, 1)
if np.min(dist) < 4e-3:
continue
shown_images = np.r_[shown_images, [X[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(digits.images[i], cmap=plt.cm.gray_r),
X[i])
ax.add_artist(imagebox)
plt.xticks([])
plt.yticks([])
if title is not None:
plt.title(title)
plt.show()
def plot_embedding(X, imgnames, title=None):
#K = 35
#rn_values = range(len(imgnames))
#shuffle(rn_values)
#sample = rn_values[:K]
#X = X[sample,:]
#imgnames = imgnames[sample]
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
#for i in range(X.shape[0]):
# plt.text(X[i, 0], X[i, 1], str(imgnames[i]),fontdict={'weight': 'bold', 'size': 9})
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
#dist = np.sum((X[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[i]]]
timg = misc.imread(imgnames[i])
timg = np.array(misc.imresize(timg, (70, 70, 3)))
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(timg, cmap=plt.cm.gray_r), X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
plt.show(block=True)
def plot_embedding(X, title=None):
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.text(X[i, 0],
X[i, 1],
str(train_target[i]),
color=plt.cm.Set1(train_target[i] / 2.),
fontdict={
'weight': 'bold',
'size': 9
})
if hasattr(offsetbox, 'AnnotationBbox'):
shown_images = np.array([[1., 1.]]) # just something big
for i in range(show_instancees):
dist = np.sum((X[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[i]]]
auctionbox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(images[train_target[i]], cmap=plt.cm.gray_r),
X[i])
ax.add_artist(auctionbox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
def acquire_teamPic(season_id, team_title, team_id, zoom, offset=(250, 370)):
abb_file = os.getcwd() + "/../csvs/team_abbreviations.csv"
abb_list = {}
with open(abb_file, 'rU') as f:
mycsv = csv.reader(f)
for row in mycsv:
team, abb, imgurl = row
abb_list[team] = [abb, imgurl]
img_url = abb_list.get(team_title)[1]
try:
img_path = os.getcwd() + '/' + str(team_id) + '.png'
team_pic = plt.imread(img_path)
except IOError:
try:
pic = urllib.urlretrieve(img_url, str(team_id) + '.png')
team_pic = plt.imread(pic[0])
except (ValueError, IOError):
img_path = os.getcwd() + '/nba_logo.png'
player_pic = plt.imread(img_path)
img = osb.OffsetImage(team_pic, zoom)
img = osb.AnnotationBbox(img,
offset,
xycoords='data',
pad=0.0,
box_alignment=(1, 0),
frameon=False)
return img
def plot_embedding(X, title, ax):
X = MinMaxScaler().fit_transform(X)
for digit in digits.target_names:
ax.scatter(
*X[y == digit].T,
marker=f"${digit}$",
s=60,
color=plt.cm.Dark2(digit),
alpha=0.425,
zorder=2,
)
shown_images = np.array([[1.0, 1.0]]) # just something big
for i in range(X.shape[0]):
# plot every digit on the embedding
# show an annotation box for a group of digits
dist = np.sum((X[i] - shown_images)**2, 1)
if np.min(dist) < 4e-3:
# don't show points that are too close
continue
shown_images = np.concatenate([shown_images, [X[i]]], axis=0)
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(digits.images[i], cmap=plt.cm.gray_r), X[i])
imagebox.set(zorder=1)
ax.add_artist(imagebox)
ax.set_title(title)
ax.axis("off")
def plot_embedding(X, title, ax):
X = MinMaxScaler().fit_transform(X)
shown_images = np.array([[1.0, 1.0]]) # just something big
for i in range(X.shape[0]):
# plot every digit on the embedding
ax.text(
X[i, 0],
X[i, 1],
str(y[i]),
color=plt.cm.Dark2(y[i]),
fontdict={"weight": "bold", "size": 9},
)
# show an annotation box for a group of digits
dist = np.sum((X[i] - shown_images) ** 2, 1)
if np.min(dist) < 4e-3:
# don't show points that are too close
continue
shown_images = np.concatenate([shown_images, [X[i]]], axis=0)
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(digits.images[i], cmap=plt.cm.gray_r), X[i]
)
ax.add_artist(imagebox)
ax.set_title(title)
ax.axis("off")
def plot_components(X_projected,
images=None,
ax=None,
image_scale=1.0,
markersize=10,
thumb_frac=0.05,
cmap='gray'):
# https://jakevdp.github.io/PythonDataScienceHandbook/05.10-manifold-learning.html
ax = ax or plt.gca()
ax.plot(X_projected[:, 0], X_projected[:, 1], '.k', markersize=markersize)
# images = images[:, ::image_scale, ::image_scale] # downsample the images
# images = imresize(images, (images.shape[0], images.shape[1]*image_scale, images.shape[2]*image_scale)) # downsample the images
images = resize(images, (images.shape[0], images.shape[1] * image_scale,
images.shape[2] * image_scale),
order=5,
preserve_range=True)
if images is not None:
min_dist_2 = (thumb_frac *
max(X_projected.max(0) - X_projected.min(0)))**2
shown_images = np.array([2 * X_projected.max(0)])
for i in range(X_projected.shape[0]):
dist = np.sum((X_projected[i] - shown_images)**2, 1)
if np.min(dist) < min_dist_2:
# don't show points that are too close
continue
shown_images = np.vstack([shown_images, X_projected[i]])
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(images[i], cmap=cmap), X_projected[i])
ax.add_artist(imagebox)
#plt.show()
#TODO: Add desired location/path + <name>.png to save plot
plt.savefig('')
def plot_embedding(X, y, title=None):
from matplotlib import offsetbox
x_min, x_max = np.min(X, 0), np.max(X, 0)
X = (X - x_min) / (x_max - x_min)
plt.figure()
ax = plt.subplot(111)
for i in range(X.shape[0]):
plt.text(X[i, 0], X[i, 1], str(y[i]),
color=plt.cm.Set1(y[i] / 10.),
fontdict={'weight': 'bold', 'size': 9})
if hasattr(offsetbox, 'AnnotationBbox'):
# only print thumbnails with matplotlib > 1.0
shown_images = np.array([[1., 1.]]) # just something big
for i in range(X.shape[0]):
dist = np.sum((X[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[i]]]
imagebox = offsetbox.AnnotationBbox(
offsetbox.OffsetImage(y[i], cmap=plt.cm.gray_r),
X[i])
ax.add_artist(imagebox)
plt.xticks([]), plt.yticks([])
if title is not None:
plt.title(title)
def visualize(embed, x_test):
# two ways of visualization: scale to fit [0,1] scale
# feat = embed - np.min(embed, 0)
# feat /= np.max(feat, 0)
# two ways of visualization: leave with original scale
feat = embed
ax_min = np.min(embed,
0) # [-11.579187 -9.2086525] 0 means col, 1 means row
ax_max = np.max(embed, 0) # [7.0608773 5.296893 ]
ax_dist_sq = np.sum((ax_max - ax_min)**2) # 557.86285
plt.figure()
ax = plt.subplot(111)
shown_images = np.array([[1., 1.]])
for i in range(feat.shape[0]):
dist = np.sum((feat[i] - shown_images)**2,
1) # 0-x axis project, 1-y axis
if np.min(
dist
) < 3e-4 * ax_dist_sq: # don't show points that are too close
continue
shown_images = np.r_[shown_images, [feat[i]]] # concatenate mat
imagebox = offsetbox.AnnotationBbox(offsetbox.OffsetImage(
x_test[i], zoom=0.6, cmap=plt.cm.gray_r),
xy=feat[i],
frameon=False)
ax.add_artist(imagebox)
plt.axis([ax_min[0], ax_max[0], ax_min[1], ax_max[1]])
# plt.xticks([]), plt.yticks([])
plt.title('Embedding from the last layer of the network')
plt.show()