Py.Cafe
Lipstick Shades t-SNE Clustering
- app.py
- requirements.txt
app.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
import dash
from dash import dcc, html
from dash.dependencies import Input, Output
import plotly.graph_objs as go
import numpy as np
import pandas as pd
from sklearn.manifold import TSNE
from sklearn.cluster import KMeans
def generate_lipstick_dataset(n=100):
np.random.seed(42)
n = n // 3 * 3
reds = np.random.randint(100, 256, (n//3, 1))
pinks = np.random.randint(50, 200, (n//3, 1))
nudes = np.random.randint(150, 230, (n//3, 1))
red_shades = np.concatenate((reds, np.random.randint(0, 100, (n//3, 2))), axis=1)
pink_shades = np.concatenate((np.random.randint(150, 256, (n//3, 1)), pinks, np.random.randint(150, 256, (n//3, 1))), axis=1)
nude_shades = np.concatenate((nudes, nudes, nudes), axis=1)
shades = np.vstack((red_shades, pink_shades, nude_shades))
df = pd.DataFrame(shades, columns=['R', 'G', 'B'])
df['Shade'] = ['Shade {}'.format(i) for i in range(n)]
return df
def perform_tsne(df, perplexity=30):
tsne = TSNE(n_components=2, perplexity=min(perplexity, len(df) - 1), random_state=42)
tsne_result = tsne.fit_transform(df[['R', 'G', 'B']])
df['TSNE-2D-One'] = tsne_result[:, 0]
df['TSNE-2D-Two'] = tsne_result[:, 1]
return df
def perform_kmeans(df, n_clusters=5):
kmeans = KMeans(n_clusters=n_clusters, random_state=42, n_init=10)
df['Cluster'] = kmeans.fit_predict(df[['TSNE-2D-One', 'TSNE-2D-Two']])
return df
def rgb_to_hex(r, g, b):
return '#{:02x}{:02x}{:02x}'.format(r, g, b)
df = generate_lipstick_dataset()
df = perform_tsne(df)
df = perform_kmeans(df)
app = dash.Dash(__name__)
# Include custom CSS in the app layout
app.index_string = '''
<!DOCTYPE html>
<html>
<head>
{%metas%}
<title>{%title%}</title>
{%favicon%}
{%css%}
<style>
body {
background-color: #F0F0F0;
font-family: 'Helvetica Neue', Helvetica, Arial, sans-serif;
margin: 0;
padding: 0;
}
h1 {
color: #1C1C1C;
text-align: center;
margin-top: 20px;
font-size: 2.5em;
font-weight: bold;
}
label {
color: #1C1C1C;
font-size: 1.2em;
margin-left: 20px;
}
.dccSlider {
width: 80%;
margin-left: auto;
margin-right: auto;
margin-top: 20px;
}
.dccGraph {
margin-left: auto;
margin-right: auto;
width: 90%;
}
.container {
max-width: 1200px;
margin: auto;
padding: 20px;
background-color: #FFFFFF;
border-radius: 10px;
box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
}
.dash-slider .rc-slider-track {
background-color: #FF0000;
}
.dash-slider .rc-slider-handle {
border-color: #FF0000;
}
.dash-slider .rc-slider-mark-text {
color: #1C1C1C;
}
</style>
</head>
<body>
{%app_entry%}
<footer>
{%config%}
{%scripts%}
{%renderer%}
</footer>
</body>
</html>
'''
app.layout = html.Div(className='container', children=[
html.H1('Lipstick Shades 2D t-SNE Clustering'),
html.Label('Number of shades:', className='dccLabel'),
dcc.Slider(
id='n-shades-slider',
min=10,
max=200,
step=10,
value=100,
marks={i: str(i) for i in range(10, 210, 20)},
className='dash-slider'
),
html.Label('Number of clusters:', className='dccLabel'),
dcc.Slider(
id='n-clusters-slider',
min=2,
max=10,
step=1,
value=5,
marks={i: str(i) for i in range(2, 11)},
className='dash-slider'
),
dcc.Graph(id='tsne-plot', className='dccGraph')
])
@app.callback(
Output('tsne-plot', 'figure'),
[Input('n-shades-slider', 'value'), Input('n-clusters-slider', 'value')]
)
def update_tsne_plot(n_shades, n_clusters):
df_sample = generate_lipstick_dataset(n=n_shades)
df_sample = perform_tsne(df_sample, perplexity=30)
df_sample = perform_kmeans(df_sample, n_clusters=n_clusters)
df_sample['hex'] = df_sample.apply(lambda row: rgb_to_hex(row['R'], row['G'], row['B']), axis=1)
df_sample['hover_text'] = df_sample.apply(
lambda row: f"Shade: {row['Shade']}<br>RGB: ({row['R']}, {row['G']}, {row['B']})<br>Hex: {row['hex']}", axis=1
)
fig = go.Figure()
fig.add_trace(go.Scatter(
x=df_sample['TSNE-2D-One'],
y=df_sample['TSNE-2D-Two'],
mode='markers',
marker=dict(
size=10,
color=['rgb({}, {}, {})'.format(r, g, b) for r, g, b in zip(df_sample['R'], df_sample['G'], df_sample['B'])]
),
text=df_sample['hover_text'],
hoverinfo='text'
))
fig.update_layout(
title=f'2D t-SNE Clustering of {n_shades} Lipstick Shades',
showlegend=False,
margin=dict(l=0, r=0, t=30, b=0),
xaxis_title='TSNE-2D-One',
yaxis_title='TSNE-2D-Two'
)
return fig
if __name__ == '__main__':
app.run_server(debug=True)