Spaces:
Sleeping
Sleeping
Commit
Β·
1d88eef
1
Parent(s):
4590034
all committed
Browse files- .gitignore +3 -0
- requirements.txt +6 -0
- tp.py +376 -0
.gitignore
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
.env
|
2 |
+
*pyc
|
3 |
+
/myenv/
|
requirements.txt
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
streamlit==1.20.0
|
2 |
+
pandas==1.5.3
|
3 |
+
plotly==5.11.0
|
4 |
+
numpy==1.24.1
|
5 |
+
astrapy==0.6.1
|
6 |
+
openai==0.27.0
|
tp.py
ADDED
@@ -0,0 +1,376 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import pandas as pd
|
3 |
+
import plotly.express as px
|
4 |
+
from astrapy import DataAPIClient
|
5 |
+
import plotly.graph_objects as go
|
6 |
+
from plotly.subplots import make_subplots
|
7 |
+
import numpy as np
|
8 |
+
from openai import OpenAI
|
9 |
+
from typing import Dict, List
|
10 |
+
from dotenv import load_dotenv
|
11 |
+
import os
|
12 |
+
|
13 |
+
# Load environment variables
|
14 |
+
load_dotenv()
|
15 |
+
|
16 |
+
def initialize_client():
|
17 |
+
try:
|
18 |
+
token = os.getenv("ASTRA_DB_TOKEN")
|
19 |
+
endpoint = os.getenv("ASTRA_DB_ENDPOINT")
|
20 |
+
|
21 |
+
if not token or not endpoint:
|
22 |
+
raise ValueError("AstraDB token or endpoint not found in environment variables.")
|
23 |
+
|
24 |
+
client = DataAPIClient(token)
|
25 |
+
db = client.get_database_by_api_endpoint(endpoint)
|
26 |
+
return db
|
27 |
+
except Exception as e:
|
28 |
+
st.error(f"Error initializing AstraDB client: {e}")
|
29 |
+
return None
|
30 |
+
|
31 |
+
def fetch_collection_data(db, collection_name):
|
32 |
+
try:
|
33 |
+
collection = db[collection_name]
|
34 |
+
documents = collection.find({})
|
35 |
+
return list(documents)
|
36 |
+
except Exception as e:
|
37 |
+
st.error(f"Error fetching data from collection {collection_name}: {e}")
|
38 |
+
return None
|
39 |
+
|
40 |
+
@st.cache_data
|
41 |
+
def process_dataframe(data):
|
42 |
+
"""Cache the dataframe processing to prevent unnecessary recomputation"""
|
43 |
+
df = pd.DataFrame(data)
|
44 |
+
df = df.apply(pd.to_numeric, errors="ignore")
|
45 |
+
return df
|
46 |
+
|
47 |
+
def create_basic_visualization(df, viz_type, x_col, y_col, color_col=None):
|
48 |
+
"""Handle basic visualization types"""
|
49 |
+
if viz_type == "Line Chart":
|
50 |
+
fig = px.line(df, x=x_col, y=y_col, color=color_col, markers=True)
|
51 |
+
elif viz_type == "Bar Chart":
|
52 |
+
fig = px.bar(df, x=x_col, y=y_col, color=color_col, text=y_col)
|
53 |
+
elif viz_type == "Scatter Plot":
|
54 |
+
fig = px.scatter(df, x=x_col, y=y_col, color=color_col, size=y_col, hover_data=[color_col])
|
55 |
+
elif viz_type == "Box Plot":
|
56 |
+
fig = px.box(df, x=x_col, y=y_col, color=color_col, points="all")
|
57 |
+
return fig
|
58 |
+
|
59 |
+
def create_advanced_visualization(df, viz_type, x_col, y_col, color_col=None):
|
60 |
+
if viz_type in ["Line Chart", "Bar Chart", "Scatter Plot", "Box Plot"]:
|
61 |
+
fig = create_basic_visualization(df, viz_type, x_col, y_col, color_col)
|
62 |
+
|
63 |
+
elif viz_type == "Engagement Sunburst":
|
64 |
+
total_engagement = df['likes'] + df['shares'] + df['comments']
|
65 |
+
engagement_labels = pd.qcut(total_engagement, q=4, labels=['Low', 'Medium', 'High', 'Viral'])
|
66 |
+
temp_df = pd.DataFrame({
|
67 |
+
'engagement_level': engagement_labels,
|
68 |
+
'post_type': df['post_type'],
|
69 |
+
'likes': df['likes'],
|
70 |
+
'sentiment': df['avg_sentiment_score']
|
71 |
+
})
|
72 |
+
|
73 |
+
fig = px.sunburst(
|
74 |
+
temp_df,
|
75 |
+
path=['engagement_level', 'post_type'],
|
76 |
+
values='likes',
|
77 |
+
color='sentiment',
|
78 |
+
color_continuous_scale='RdYlBu',
|
79 |
+
title="Engagement Distribution by Post Type and Sentiment"
|
80 |
+
)
|
81 |
+
|
82 |
+
elif viz_type == "Sentiment Heat Calendar":
|
83 |
+
# Create dummy datetime for visualization
|
84 |
+
hour_data = []
|
85 |
+
days = ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']
|
86 |
+
|
87 |
+
for day in days:
|
88 |
+
for hour in range(24):
|
89 |
+
avg_sentiment = df['avg_sentiment_score'].mean() + np.random.normal(0, 0.1)
|
90 |
+
hour_data.append({
|
91 |
+
'day': day,
|
92 |
+
'hour': hour,
|
93 |
+
'sentiment': avg_sentiment
|
94 |
+
})
|
95 |
+
|
96 |
+
temp_df = pd.DataFrame(hour_data)
|
97 |
+
fig = px.density_heatmap(
|
98 |
+
temp_df,
|
99 |
+
x='day',
|
100 |
+
y='hour',
|
101 |
+
z='sentiment',
|
102 |
+
title="Sentiment Distribution by Day and Hour",
|
103 |
+
labels={'sentiment': 'Average Sentiment'},
|
104 |
+
color_continuous_scale="RdYlBu"
|
105 |
+
)
|
106 |
+
|
107 |
+
elif viz_type == "Engagement Spider":
|
108 |
+
metrics = ['likes', 'shares', 'comments']
|
109 |
+
df_norm = df[metrics].apply(lambda x: (x - x.min()) / (x.max() - x.min()))
|
110 |
+
|
111 |
+
fig = go.Figure()
|
112 |
+
for ptype in df['post_type'].unique():
|
113 |
+
values = df_norm[df['post_type'] == ptype].mean()
|
114 |
+
fig.add_trace(go.Scatterpolar(
|
115 |
+
r=values.tolist() + [values.iloc[0]],
|
116 |
+
theta=metrics + [metrics[0]],
|
117 |
+
name=ptype,
|
118 |
+
fill='toself'
|
119 |
+
))
|
120 |
+
|
121 |
+
fig.update_layout(
|
122 |
+
polar=dict(radialaxis=dict(visible=True, range=[0, 1])),
|
123 |
+
showlegend=True,
|
124 |
+
title="Engagement Pattern by Post Type"
|
125 |
+
)
|
126 |
+
|
127 |
+
elif viz_type == "Sentiment Flow":
|
128 |
+
# Group by post type and calculate rolling average
|
129 |
+
fig = go.Figure()
|
130 |
+
for ptype in df['post_type'].unique():
|
131 |
+
mask = df['post_type'] == ptype
|
132 |
+
sentiment_series = df[mask]['avg_sentiment_score']
|
133 |
+
rolling_avg = sentiment_series.rolling(window=min(7, len(sentiment_series))).mean()
|
134 |
+
|
135 |
+
fig.add_trace(go.Scatter(
|
136 |
+
x=list(range(len(rolling_avg))), # Use index instead of dates
|
137 |
+
y=rolling_avg,
|
138 |
+
name=ptype,
|
139 |
+
mode='lines',
|
140 |
+
fill='tonexty'
|
141 |
+
))
|
142 |
+
|
143 |
+
fig.update_layout(
|
144 |
+
title="Sentiment Flow by Post Type",
|
145 |
+
xaxis_title="Post Sequence",
|
146 |
+
yaxis_title="Average Sentiment"
|
147 |
+
)
|
148 |
+
|
149 |
+
elif viz_type == "Engagement Matrix":
|
150 |
+
corr_matrix = df[['likes', 'shares', 'comments', 'avg_sentiment_score']].corr()
|
151 |
+
|
152 |
+
fig = px.imshow(
|
153 |
+
corr_matrix,
|
154 |
+
color_continuous_scale='RdBu',
|
155 |
+
aspect='auto',
|
156 |
+
title="Engagement Metrics Correlation Matrix"
|
157 |
+
)
|
158 |
+
|
159 |
+
# Apply theme
|
160 |
+
fig.update_layout(
|
161 |
+
template="plotly_dark" if st.session_state.dark_mode else "plotly_white",
|
162 |
+
title_x=0.5,
|
163 |
+
font=dict(size=14),
|
164 |
+
margin=dict(l=20, r=20, t=50, b=20),
|
165 |
+
paper_bgcolor="#1e1e1e" if st.session_state.dark_mode else "#f9f9f9",
|
166 |
+
plot_bgcolor="#1e1e1e" if st.session_state.dark_mode else "#f9f9f9",
|
167 |
+
)
|
168 |
+
return fig
|
169 |
+
|
170 |
+
def initialize_openai():
|
171 |
+
"""Initialize OpenAI client"""
|
172 |
+
try:
|
173 |
+
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
|
174 |
+
return client
|
175 |
+
except Exception as e:
|
176 |
+
st.error(f"Error initializing OpenAI: {e}")
|
177 |
+
return None
|
178 |
+
|
179 |
+
def generate_prompt(metrics: Dict) -> str:
|
180 |
+
"""Generate a prompt for GPT based on the metrics"""
|
181 |
+
return f"""Analyze the following social media metrics and provide 3-5 clear, specific insights about post performance:
|
182 |
+
|
183 |
+
Post Type Metrics:
|
184 |
+
{metrics}
|
185 |
+
|
186 |
+
Please focus on:
|
187 |
+
1. Comparative performance between post types
|
188 |
+
2. Engagement patterns
|
189 |
+
3. Notable trends or anomalies
|
190 |
+
4. Actionable recommendations
|
191 |
+
|
192 |
+
Format your response in clear bullet points with percentage comparisons where relevant.
|
193 |
+
Keep each insight concise but specific, including numerical comparisons.
|
194 |
+
"""
|
195 |
+
|
196 |
+
def calculate_metrics(df: pd.DataFrame) -> Dict:
|
197 |
+
"""Calculate comprehensive metrics for GPT analysis"""
|
198 |
+
metrics = {}
|
199 |
+
|
200 |
+
# Calculate per post type metrics
|
201 |
+
for post_type in df['post_type'].unique():
|
202 |
+
post_data = df[df['post_type'] == post_type]
|
203 |
+
metrics[post_type] = {
|
204 |
+
'avg_likes': post_data['likes'].mean(),
|
205 |
+
'avg_shares': post_data['shares'].mean(),
|
206 |
+
'avg_comments': post_data['comments'].mean(),
|
207 |
+
'avg_sentiment': post_data['avg_sentiment_score'].mean(),
|
208 |
+
'engagement_rate': (post_data['likes'] + post_data['shares'] + post_data['comments']).mean(),
|
209 |
+
'post_count': len(post_data)
|
210 |
+
}
|
211 |
+
|
212 |
+
# Calculate comparative metrics
|
213 |
+
total_posts = len(df)
|
214 |
+
total_engagement = df['likes'].sum() + df['shares'].sum() + df['comments'].sum()
|
215 |
+
|
216 |
+
metrics['overall'] = {
|
217 |
+
'total_posts': total_posts,
|
218 |
+
'total_engagement': total_engagement,
|
219 |
+
'avg_sentiment_overall': df['avg_sentiment_score'].mean()
|
220 |
+
}
|
221 |
+
|
222 |
+
return metrics
|
223 |
+
|
224 |
+
def get_gpt_insights(client: OpenAI, metrics: Dict, user_query: str) -> str:
|
225 |
+
"""Get insights from GPT based on the metrics and user query"""
|
226 |
+
try:
|
227 |
+
prompt = generate_prompt(metrics) + f"\n\nUser Query: {user_query}"
|
228 |
+
|
229 |
+
response = client.chat.completions.create(
|
230 |
+
model="gpt-3.5-turbo",
|
231 |
+
messages=[
|
232 |
+
{"role": "system", "content": "You are a social media analytics expert. Provide clear, specific insights based on the data."},
|
233 |
+
{"role": "user", "content": prompt}
|
234 |
+
],
|
235 |
+
temperature=0.7,
|
236 |
+
max_tokens=500
|
237 |
+
)
|
238 |
+
|
239 |
+
# Extract and clean insights
|
240 |
+
insights_text = response.choices[0].message.content
|
241 |
+
return insights_text.strip()
|
242 |
+
|
243 |
+
except Exception as e:
|
244 |
+
return f"Error generating insights: {e}"
|
245 |
+
|
246 |
+
def main():
|
247 |
+
st.set_page_config(
|
248 |
+
page_title="Advanced Social Media Analytics Dashboard",
|
249 |
+
page_icon="π",
|
250 |
+
layout="wide",
|
251 |
+
)
|
252 |
+
openai_client = initialize_openai()
|
253 |
+
|
254 |
+
# Sidebar Settings
|
255 |
+
with st.sidebar:
|
256 |
+
st.title("Dashboard Settings")
|
257 |
+
if "dark_mode" not in st.session_state:
|
258 |
+
st.session_state.dark_mode = False
|
259 |
+
st.checkbox("Dark Mode", value=st.session_state.dark_mode, key="dark_mode")
|
260 |
+
|
261 |
+
st.write("### Data Source")
|
262 |
+
st.info("Initializing connection to AstraDB...")
|
263 |
+
db = initialize_client()
|
264 |
+
if not db:
|
265 |
+
return
|
266 |
+
|
267 |
+
collections = db.list_collection_names()
|
268 |
+
st.success("Connected to AstraDB")
|
269 |
+
selected_collection = st.selectbox("Select Collection", collections)
|
270 |
+
|
271 |
+
if selected_collection:
|
272 |
+
data = fetch_collection_data(db, selected_collection)
|
273 |
+
if data:
|
274 |
+
# Use cached data processing
|
275 |
+
df = process_dataframe(data)
|
276 |
+
|
277 |
+
# Create tabs for different analysis views
|
278 |
+
tab1, tab2, tab3 = st.tabs(["π Visualizations", "π Metrics", "π€ AI Insights"])
|
279 |
+
with tab1:
|
280 |
+
col1, col2 = st.columns([1, 3])
|
281 |
+
|
282 |
+
with col1:
|
283 |
+
st.write("### Visualization Options")
|
284 |
+
viz_type = st.selectbox(
|
285 |
+
"Select Analysis Type",
|
286 |
+
[
|
287 |
+
"Engagement Sunburst",
|
288 |
+
"Sentiment Heat Calendar",
|
289 |
+
"Engagement Spider",
|
290 |
+
"Sentiment Flow",
|
291 |
+
"Engagement Matrix",
|
292 |
+
"Line Chart",
|
293 |
+
"Bar Chart",
|
294 |
+
"Scatter Plot",
|
295 |
+
"Box Plot"
|
296 |
+
]
|
297 |
+
)
|
298 |
+
|
299 |
+
if viz_type in ["Line Chart", "Bar Chart", "Scatter Plot", "Box Plot"]:
|
300 |
+
x_col = st.selectbox("Select X-axis", df.columns)
|
301 |
+
y_col = st.selectbox("Select Y-axis", df.select_dtypes(include=["number"]).columns)
|
302 |
+
color_col = st.selectbox("Select Color Column (Optional)", [None] + list(df.columns), index=0)
|
303 |
+
else:
|
304 |
+
x_col = y_col = color_col = None
|
305 |
+
|
306 |
+
with col2:
|
307 |
+
try:
|
308 |
+
fig = create_advanced_visualization(df, viz_type, x_col, y_col, color_col)
|
309 |
+
st.plotly_chart(fig, use_container_width=True)
|
310 |
+
except Exception as e:
|
311 |
+
st.error(f"Error creating visualization: {e}")
|
312 |
+
|
313 |
+
with tab2:
|
314 |
+
# Display key metrics and insights
|
315 |
+
col1, col2, col3 = st.columns(3)
|
316 |
+
|
317 |
+
with col1:
|
318 |
+
st.metric("Average Engagement Rate",
|
319 |
+
f"{((df['likes'] + df['shares'] + df['comments']).mean() / len(df)):.2f}")
|
320 |
+
st.metric("Likes Mean", f"{df['likes'].mean():.2f}")
|
321 |
+
st.metric("Shares Mean", f"{df['shares'].mean():.2f}")
|
322 |
+
st.metric("Comments Mean", f"{df['comments'].mean():.2f}")
|
323 |
+
st.metric("Max Likes", f"{df['likes'].max():.2f}")
|
324 |
+
st.metric("Min Likes", f"{df['likes'].min():.2f}")
|
325 |
+
|
326 |
+
with col2:
|
327 |
+
st.metric("Sentiment Trend",
|
328 |
+
f"{df['avg_sentiment_score'].mean():.2f}",
|
329 |
+
f"{df['avg_sentiment_score'].std():.2f}")
|
330 |
+
st.metric("Max Shares", f"{df['shares'].max():.2f}")
|
331 |
+
st.metric("Min Shares", f"{df['shares'].min():.2f}")
|
332 |
+
st.metric("Max Comments", f"{df['comments'].max():.2f}")
|
333 |
+
st.metric("Min Comments", f"{df['comments'].min():.2f}")
|
334 |
+
st.metric("Median Sentiment", f"{df['avg_sentiment_score'].median():.2f}")
|
335 |
+
|
336 |
+
with col3:
|
337 |
+
top_type = df.groupby('post_type')['likes'].sum().idxmax()
|
338 |
+
st.metric("Most Engaging Post Type", top_type)
|
339 |
+
|
340 |
+
with st.expander("Detailed Post Overview"):
|
341 |
+
st.markdown("**Detailed metrics for each post (ID, likes, shares, comments, sentiment):**")
|
342 |
+
if 'post_id' in df.columns:
|
343 |
+
st.dataframe(df[['post_id','likes','shares','comments','avg_sentiment_score']])
|
344 |
+
else:
|
345 |
+
st.warning("No 'post_id' column found in the data.")
|
346 |
+
|
347 |
+
with tab3:
|
348 |
+
st.write("## AI Chatbot Insights")
|
349 |
+
if not openai_client:
|
350 |
+
st.error("OpenAI API not configured. Please add your API key to access AI insights.")
|
351 |
+
else:
|
352 |
+
if 'chat_history' not in st.session_state:
|
353 |
+
st.session_state.chat_history = []
|
354 |
+
|
355 |
+
user_input = st.text_input("Ask about data or insights:")
|
356 |
+
if st.button("Send"):
|
357 |
+
st.session_state.chat_history.append({"role": "user", "content": user_input})
|
358 |
+
|
359 |
+
# Use the modified get_gpt_insights function to generate response
|
360 |
+
metrics = calculate_metrics(df)
|
361 |
+
reply = get_gpt_insights(openai_client, metrics, user_input)
|
362 |
+
st.session_state.chat_history.append({"role": "assistant", "content": reply})
|
363 |
+
|
364 |
+
for msg in st.session_state.chat_history:
|
365 |
+
if msg["role"] == "user":
|
366 |
+
st.markdown(f"**You:** {msg['content']}")
|
367 |
+
else:
|
368 |
+
st.markdown(f"**Assistant:** {msg['content']}")
|
369 |
+
|
370 |
+
else:
|
371 |
+
st.error("Failed to fetch data from the selected collection.")
|
372 |
+
else:
|
373 |
+
st.error("Please select a valid collection.")
|
374 |
+
|
375 |
+
if __name__ == "__main__":
|
376 |
+
main()
|