-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathRecommendation.py
359 lines (277 loc) · 10.8 KB
/
Recommendation.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
####Data Preparation####
from mpl_toolkits.mplot3d import Axes3D
from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt # plotting
import matplotlib.image as mpimg
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import os # accessing directory structure
DATASET_PATH = "/content/drive/MyDrive/test/data/fashion4/" # specify folder path
print(os.listdir(DATASET_PATH)) # show file names in folder
df = pd.read_csv(DATASET_PATH + "styles.csv", nrows=5000, error_bad_lines=False) # read the csv file from the specified path as a dataframe
df['image'] = df.apply(lambda row: str(row['id']) + ".jpg", axis=1) # add a new column based on the id column and name it image
df = df.reset_index(drop=True) # rearrange index
df.head(10) # display the first ten rows of the dataframe
import cv2
def plot_figures(figures, nrows = 1, ncols = 1, figsize = (8, 8)):
"""Plot a dictionary of figures
Parameters
----------
figures : <title, figure> dictionary
ncols : number of columns of subplots wanted in the display
nrows : number of rows of subplots wanted in the figure
"""
fig, axeslist = plt.subplots(ncols=ncols, nrows=nrows,figsize=figsize)
for ind,title in enumerate(figures):
axeslist.ravel()[ind].imshow(cv2.cvtColor(figures[title], cv2.COLOR_BGR2RGB))
axeslist.ravel()[ind].set_title(title)
axeslist.ravel()[ind].set_axis_off()
plt.tight_layout() # optional
def img_path(img):
"""Get image path
Parameters
----------
img : image name
"""
return DATASET_PATH+"images/"+img
def load_image(img, resized_fac = 3):
"""Get image path
Parameters
----------
img : image name
resized_fac ： the size fraction you want to resized
"""
img = cv2.imread(img_path(img)) # read image
if type(img) == np.ndarray:
w, h, _ = img.shape # view image size
resized = cv2.resize(img, (int(h*resized_fac), int(w*resized_fac)), interpolation = cv2.INTER_LINEAR) # adjust pixel values using linear interpolation
return resized
else:
return img
# search images that can not be read
result = df['image'].apply(load_image)
delt = [i for i in range(len(result)) if type(result[i]) != np.ndarray]
delt
df = df.drop(delt) # delete images that can not be read
df = df.reset_index(drop=True) # rearrange index
import matplotlib.pyplot as plt
import numpy as np
# generation of a dictionary of (title, images)
figures = {'im'+str(i): load_image(row.image) for i, row in df.sample(6).iterrows()}
# plot of the images in a figure, with 2 rows and 3 columns
plot_figures(figures, 2, 3)
plt.figure(figsize=(7,20))
df.articleType.value_counts().sort_values().plot(kind='barh')
####Use Pre-Trained Model to Recommendation####
import tensorflow as tf
import keras
from keras import Model
from tensorflow.keras.applications.resnet50 import ResNet50
from keras.preprocessing import image
from tensorflow.keras.applications.resnet50 import preprocess_input, decode_predictions
from keras.layers import GlobalMaxPooling2D
tf.__version__
# Input Shape
img_width, img_height, _ = 224, 224, 3 # load_image(df.iloc[0].image).shape
# Pre-Trained Model
base_model = ResNet50(weights='imagenet', include_top=False, input_shape = (img_width, img_height, 3))
base_model.trainable = False
# Add Layer Embedding
model = keras.Sequential([base_model, GlobalMaxPooling2D()])
model.summary()
def get_embedding1(model, img_name):
# reshape
img = tf.keras.utils.load_img(img_path(img_name), target_size=(img_width, img_height))
# img to array
x = tf.keras.utils.img_to_array(img)
# expand dim (1, w, h)
x = np.expand_dims(x, axis=0)
# pre process input
x = preprocess_input(x)
return model.predict(x).reshape(-1)
##Get item Embedding
emb = get_embedding1(model, df.iloc[0].image)
emb.shape
img_array = load_image(df.iloc[0].image)
plt.imshow(cv2.cvtColor(img_array, cv2.COLOR_BGR2RGB))
print(img_array.shape)
print(emb)
df.shape
##Get Embedding for all itens in dataset
# import swifter
# Parallel Apply
df_sample = df#.sample(10)
map_embeddings1 = df_sample['image'].apply(lambda img: get_embedding1(model, img))
df_embs1 = map_embeddings1.apply(pd.Series)
print(df_embs1.shape)
df_embs1.head()
##Compute Similarity Between Items
from sklearn.metrics.pairwise import pairwise_distances
# Calcule Distance Matrix
cosine_sim1 = 1 - pairwise_distances(df_embs1, metric='cosine')
cosine_sim1[:4, :4]
##Recommender Similar Items
indices = pd.Series(range(len(df)), index=df.index)
indices
# function that get movie recommendations based on the cosine similarity score of movie genres
def get_recommender1(idx, df, top_n = 5):
sim_idx = indices[idx]
sim_scores = list(enumerate(cosine_sim1[sim_idx]))
sim_scores = sorted(sim_scores, key=lambda x: x[1], reverse=True)
sim_scores = sim_scores[1:top_n+1]
idx_rec = [i[0] for i in sim_scores]
idx_sim = [i[1] for i in sim_scores]
return indices.iloc[idx_rec].index, idx_sim
get_recommender1(2993, df, top_n = 5)
idx_ref = 0
# Recommendations
idx_rec, idx_sim = get_recommender1(idx_ref, df, top_n = 6)
# Plot
plt.imshow(cv2.cvtColor(load_image(df.iloc[idx_ref].image), cv2.COLOR_BGR2RGB))
# generation of a dictionary of (title, images)
figures = {'im'+str(i): load_image(row.image) for i, row in df.loc[idx_rec].iterrows()}
# plot of the images in a figure, with 2 rows and 3 columns
plot_figures(figures, 2, 3)
####Use Convolutional Autoencoder to Recommendation####
img = tf.keras.utils.load_img(img_path(df['image'].iloc[0]), target_size=(28, 28))
# img to array
x = tf.keras.utils.img_to_array(img)
# expand dim (1, w, h)
x = np.expand_dims(x, axis=0)
x.shape
def get_x(img_name):
# reshape
img = tf.keras.utils.load_img(img_path(img_name), target_size=(28, 28))
# img to array
x = tf.keras.utils.img_to_array(img)
# expand dim (1, w, h)
x = np.expand_dims(x, axis=0)
# pre process input
x = preprocess_input(x)
x = x.astype('float32')/255.
return x
df_all = df
df_train = df.sample(frac=0.9, random_state=1234, axis=0)
df_val = df.drop(df_train.index).sample(frac=0.5, random_state=1234, axis=0)
df_test = df.drop(df_train.index).drop(df_val.index)
df_all = df_all['image'].apply(lambda img: get_x(img))
df_train = df_train['image'].apply(lambda img: get_x(img))
df_val = df_val['image'].apply(lambda img: get_x(img))
df_test = df_test['image'].apply(lambda img: get_x(img))
x_all = np.concatenate(df_all.to_list(), axis=0)
x_train = np.concatenate(df_train.to_list(), axis=0)
x_val = np.concatenate(df_val.to_list(), axis=0)
x_test = np.concatenate(df_test.to_list(), axis=0)
import keras
from keras import layers
input_img = keras.Input(shape=(28, 28, 3))
x = layers.Conv2D(16, (3, 3), activation='relu', padding='same')(input_img)
x = layers.MaxPooling2D((2, 2), padding='same')(x)
x = layers.Conv2D(8, (3, 3), activation='relu', padding='same')(x)
x = layers.MaxPooling2D((2, 2), padding='same')(x)
x = layers.Conv2D(8, (3, 3), activation='relu', padding='same')(x)
encoded = layers.MaxPooling2D((2, 2), padding='same')(x)
# at this point the representation is (4, 4, 8) i.e. 128-dimensional
x = layers.Conv2D(8, (3, 3), activation='relu', padding='same')(encoded)
x = layers.UpSampling2D((2, 2))(x)
x = layers.Conv2D(8, (3, 3), activation='relu', padding='same')(x)
x = layers.UpSampling2D((2, 2))(x)
x = layers.Conv2D(16, (3, 3), activation='relu')(x)
x = layers.UpSampling2D((2, 2))(x)
decoded = layers.Conv2D(3, (3, 3), activation='sigmoid', padding='same')(x)
autoencoder = keras.Model(input_img, decoded)
autoencoder.compile(optimizer='adam', loss='mse')
autoencoder.summary()
autoencoder.fit(x_train, x_train,
epochs = 50,
batch_size = 128,
shuffle = True,
validation_data = (x_val, x_val))
decoded_imgs = autoencoder.predict(x_test)
decoded_imgs.shape
encoder = keras.Model(input_img, encoded)
encoded_img = encoder.predict(x_test)
encoded_img.shape
def get_embedding2(x):
# flatten
x = x.flatten()
# pre process input
x = preprocess_input(x)
return x
emb2 = encoder.predict(x_all)
emb2.shape
df_embs2 = [get_embedding2(x) for x in emb2]
df_embs2 = pd.DataFrame(df_embs2)
df_embs2.head()
##Compute Similarity Between Items
from sklearn.metrics.pairwise import pairwise_distances
# Calcule Distance Matrix
cosine_sim2 = 1 - pairwise_distances(df_embs2, metric='cosine')
cosine_sim2[:4, :4]
##Recommender Similar Items
indices = pd.Series(range(len(df)), index=df.index)
indices
def get_recommender2(idx, df, top_n = 5):
sim_idx = indices[idx]
sim_scores = list(enumerate(cosine_sim2[sim_idx]))
sim_scores = sorted(sim_scores, key=lambda x: x[1], reverse=True)
sim_scores = sim_scores[1:top_n+1]
idx_rec = [i[0] for i in sim_scores]
idx_sim = [i[1] for i in sim_scores]
return indices.iloc[idx_rec].index, idx_sim
get_recommender2(2993, df, top_n = 5)
##Recommender Similar Items
# Idx Item to Recommender
idx_ref = 0
# Recommendations
idx_rec, idx_sim = get_recommender2(idx_ref, df, top_n = 6)
# Plot
plt.imshow(cv2.cvtColor(load_image(df.iloc[idx_ref].image), cv2.COLOR_BGR2RGB))
# generation of a dictionary of (title, images)
figures = {'im'+str(i): load_image(row.image) for i, row in df.loc[idx_rec].iterrows()}
# plot of the images in a figure, with 2 rows and 3 columns
plot_figures(figures, 2, 3)
##Visualization Latent Space of Contents
from sklearn.manifold import TSNE
import time
import seaborn as sns
df.shape
df.head()
df_embs1.shape
df_embs2.shape
time_start = time.time()
tsne = TSNE(n_components=2, verbose=0, perplexity=40, n_iter=300)
tsne_results1 = tsne.fit_transform(df_embs1)
print('t-SNE done! Time elapsed: {} seconds'.format(time.time()-time_start))
df['tsne-2d-one-1'] = tsne_results1[:,0]
df['tsne-2d-two-1'] = tsne_results1[:,1]
time_start = time.time()
tsne = TSNE(n_components=2, verbose=0, perplexity=40, n_iter=300)
tsne_results2 = tsne.fit_transform(df_embs2)
print('t-SNE done! Time elapsed: {} seconds'.format(time.time()-time_start))
df['tsne-2d-one-2'] = tsne_results2[:,0]
df['tsne-2d-two-2'] = tsne_results2[:,1]
plt.figure(figsize=(16,10))
sns.scatterplot(x="tsne-2d-one-1", y="tsne-2d-two-1",
hue="masterCategory",
data=df,
legend="full",
alpha=0.8)
plt.figure(figsize=(16,10))
sns.scatterplot(x="tsne-2d-one-2", y="tsne-2d-two-2",
hue="masterCategory",
data=df,
legend="full",
alpha=0.8)
plt.figure(figsize=(16,10))
sns.scatterplot(x="tsne-2d-one-1", y="tsne-2d-two-1",
hue="subCategory",
data=df,
legend="full",
alpha=0.8)
plt.figure(figsize=(16,10))
sns.scatterplot(x="tsne-2d-one-2", y="tsne-2d-two-2",
hue="subCategory",
data=df,
legend="full",
alpha=0.8)