Spaces:
Sleeping
Sleeping
Update src/streamlit_app.py
Browse files- src/streamlit_app.py +8 -5
src/streamlit_app.py
CHANGED
|
@@ -1,3 +1,4 @@
|
|
|
|
|
| 1 |
import os
|
| 2 |
os.environ['HF_HOME'] = '/tmp'
|
| 3 |
import time
|
|
@@ -395,14 +396,17 @@ def generate_html_report(df, text_input, elapsed_time, df_topic_data):
|
|
| 395 |
<p><strong>Generated At:</strong> {time.strftime('%Y-%m-%d %H:%M:%S')}</p>
|
| 396 |
<p><strong>Processing Time:</strong> {elapsed_time:.2f} seconds</p>
|
| 397 |
</div>
|
|
|
|
| 398 |
<h2>1. Analyzed Text & Extracted Entities</h2>
|
| 399 |
<h3>Original Text with Highlighted Entities</h3>
|
| 400 |
<div class="highlighted-text-container">
|
| 401 |
{highlighted_text_html}
|
| 402 |
</div>
|
| 403 |
|
|
|
|
| 404 |
<h2>2. Full Extracted Entities Table</h2>
|
| 405 |
{entity_table_html}
|
|
|
|
| 406 |
<h2>3. Data Visualizations</h2>
|
| 407 |
|
| 408 |
<h3>3.1 Entity Distribution Treemap</h3>
|
|
@@ -416,6 +420,7 @@ def generate_html_report(df, text_input, elapsed_time, df_topic_data):
|
|
| 416 |
<h3>3.3 Entity Co-occurrence Network (Edges = Same Sentence)</h3>
|
| 417 |
<div class="chart-box">{network_html}</div>
|
| 418 |
|
|
|
|
| 419 |
<h2>4. Topic Modeling (LDA on Entities)</h2>
|
| 420 |
{topic_charts_html}
|
| 421 |
|
|
@@ -577,17 +582,18 @@ if st.button("Results"):
|
|
| 577 |
st.session_state.topic_results = perform_topic_modeling(
|
| 578 |
df,
|
| 579 |
num_topics=2,
|
| 580 |
-
|
| 581 |
)
|
| 582 |
|
| 583 |
if comet_initialized:
|
|
|
|
| 584 |
experiment = Experiment(api_key=COMET_API_KEY, workspace=COMET_WORKSPACE, project_name=COMET_PROJECT_NAME)
|
| 585 |
experiment.log_parameter("input_text", text)
|
| 586 |
experiment.log_table("predicted_entities", df)
|
| 587 |
experiment.end()
|
| 588 |
else:
|
| 589 |
st.session_state.results_df = pd.DataFrame()
|
| 590 |
-
|
| 591 |
|
| 592 |
end_time = time.time()
|
| 593 |
st.session_state.elapsed_time = end_time - start_time
|
|
@@ -740,6 +746,3 @@ if st.session_state.show_results:
|
|
| 740 |
file_name=f"entity_topic_report_{time.strftime('%Y%m%d_%H%M%S')}.html",
|
| 741 |
mime="text/html",
|
| 742 |
)
|
| 743 |
-
|
| 744 |
-
|
| 745 |
-
|
|
|
|
| 1 |
+
|
| 2 |
import os
|
| 3 |
os.environ['HF_HOME'] = '/tmp'
|
| 4 |
import time
|
|
|
|
| 396 |
<p><strong>Generated At:</strong> {time.strftime('%Y-%m-%d %H:%M:%S')}</p>
|
| 397 |
<p><strong>Processing Time:</strong> {elapsed_time:.2f} seconds</p>
|
| 398 |
</div>
|
| 399 |
+
<!-- Section 1: Original Text & Highlighted Entities -->
|
| 400 |
<h2>1. Analyzed Text & Extracted Entities</h2>
|
| 401 |
<h3>Original Text with Highlighted Entities</h3>
|
| 402 |
<div class="highlighted-text-container">
|
| 403 |
{highlighted_text_html}
|
| 404 |
</div>
|
| 405 |
|
| 406 |
+
<!-- Section 2: Full Extracted Entities Table -->
|
| 407 |
<h2>2. Full Extracted Entities Table</h2>
|
| 408 |
{entity_table_html}
|
| 409 |
+
<!-- Section 3: Visualizations (Treemap, Pie, Bar Charts) -->
|
| 410 |
<h2>3. Data Visualizations</h2>
|
| 411 |
|
| 412 |
<h3>3.1 Entity Distribution Treemap</h3>
|
|
|
|
| 420 |
<h3>3.3 Entity Co-occurrence Network (Edges = Same Sentence)</h3>
|
| 421 |
<div class="chart-box">{network_html}</div>
|
| 422 |
|
| 423 |
+
<!-- Section 4: Topic Modeling -->
|
| 424 |
<h2>4. Topic Modeling (LDA on Entities)</h2>
|
| 425 |
{topic_charts_html}
|
| 426 |
|
|
|
|
| 582 |
st.session_state.topic_results = perform_topic_modeling(
|
| 583 |
df,
|
| 584 |
num_topics=2,
|
| 585 |
+
num_top_words=N_TOP_WORDS_TO_USE
|
| 586 |
)
|
| 587 |
|
| 588 |
if comet_initialized:
|
| 589 |
+
# FIX APPLIED HERE: Corrected indentation for the following lines
|
| 590 |
experiment = Experiment(api_key=COMET_API_KEY, workspace=COMET_WORKSPACE, project_name=COMET_PROJECT_NAME)
|
| 591 |
experiment.log_parameter("input_text", text)
|
| 592 |
experiment.log_table("predicted_entities", df)
|
| 593 |
experiment.end()
|
| 594 |
else:
|
| 595 |
st.session_state.results_df = pd.DataFrame()
|
| 596 |
+
st.session_state.topic_results = None
|
| 597 |
|
| 598 |
end_time = time.time()
|
| 599 |
st.session_state.elapsed_time = end_time - start_time
|
|
|
|
| 746 |
file_name=f"entity_topic_report_{time.strftime('%Y%m%d_%H%M%S')}.html",
|
| 747 |
mime="text/html",
|
| 748 |
)
|
|
|
|
|
|
|
|
|