Update app.py

app.py

@@ -1,6 +1,5 @@
 import streamlit as st
 import pandas as pd
-import plotly.express as px
 from transformers import pipeline
 
 # Set the page layout for Streamlit
@@ -12,24 +11,17 @@ tqa = pipeline(task="table-question-answering",
               device=0)  # Assuming GPU is available, otherwise set device="cpu"
 
 # Title and Introduction
-st.title("Data Visualization App with TAPAS NLP Integration")
+st.title("Data Table with TAPAS NLP Integration")
 st.markdown(""" 
-    This app allows you to upload a table (CSV or Excel) and ask questions to generate graphs visualizing the data.
-    Using **TAPAS**, the app can interpret your questions and generate the corresponding graphs.
+    This app allows you to upload a table (CSV or Excel) and ask questions to extract information from the data.
+    Using **TAPAS**, the app can interpret your questions and provide the corresponding answers.
 
     ### Available Features:
-    - **Scatter Plot**: Visualize relationships between two columns.
-    - **Line Graph**: Visualize a single column over time.
+    - **Table Question Answering**: Ask questions related to the uploaded table.
     
-    Upload your data and ask questions about the data to generate visualizations.
+    Upload your data and ask questions to extract answers.
 """)
 
-# Language Selection
-language = st.selectbox(
-    "Select the language of your question",
-    ("English", "German", "French", "Spanish", "Italian", "Others")
-)
-
 # File uploader in the sidebar
 file_name = st.sidebar.file_uploader("Upload file:", type=['csv', 'xlsx'])
 
@@ -48,81 +40,46 @@ else:
             df = None
 
         if df is not None:
-            # Show the original data with text columns intact
+            # Convert object columns to numeric where possible
+            df = df.apply(pd.to_numeric, errors='ignore')
+
             st.write("Original Data:")
             st.write(df)
 
-            # Display a sample of data for graph generation
-            st.write("Sample data for graph generation:")
+            # Display a sample of data for user reference
+            st.write("Sample data:")
             st.write(df.head())
 
     except Exception as e:
         st.error(f"Error reading file: {str(e)}")
 
     # User input for the question
-    question = st.text_input(f'Ask your graph-related question in {language}')
-
-   with st.spinner():
-    if st.button('Generate Graph'):
-        try:
-            # Ensure the question is a valid string
-            if not question or not isinstance(question, str):
-                st.error("Please enter a valid question in the form of text.")
-            else:
-                # Use TAPAS model to process the question
-                result = tqa(table=df, query=question)
-
-                # Display the raw output from TAPAS
-                st.write("TAPAS Raw Output (Response):")
-                st.write(result)  # This will display the raw output from TAPAS
-
-                # Optionally, you can output the raw output as plain text:
-                st.text("Raw TAPAS Output (Plain Text):")
-                st.text(str(result))  # This will display raw output as plain text
-
-                # Check if TAPAS is returning the expected answer
-                answer = result.get('answer', None)
-                if answer:
-                    st.write(f"TAPAS Answer: {answer}")
+    question = st.text_input(f'Ask your question related to the table')
+
+    with st.spinner():
+        if st.button('Get Answer'):
+            try:
+                # Ensure the question is a valid string
+                if not question or not isinstance(question, str):
+                    st.error("Please enter a valid question.")
                 else:
-                    st.warning("TAPAS did not return a valid answer.")
-
-                # Determine if the question relates to graph generation
-                if 'between' in question.lower() and 'and' in question.lower():
-                    # This is a request for a scatter plot (two columns)
-                    columns = question.split('between')[-1].split('and')
-                    columns = [col.strip() for col in columns]
-                    if len(columns) == 2 and all(col in df.columns for col in columns):
-                        # Prepare the data for Plotly (scatter plot)
-                        x_data = df[columns[0]].dropna()  # Extract x column, drop NaN values
-                        y_data = df[columns[1]].dropna()  # Extract y column, drop NaN values
-
-                        # Ensure x_data and y_data have the same length
-                        min_length = min(len(x_data), len(y_data))
-                        x_data = x_data[:min_length]
-                        y_data = y_data[:min_length]
-
-                        # Create the scatter plot
-                        fig = px.scatter(x=x_data, y=y_data, title=f"Scatter Plot between {columns[0]} and {columns[1]}")
-                        st.plotly_chart(fig, use_container_width=True)
-                        st.success(f"Here is the scatter plot between '{columns[0]}' and '{columns[1]}'.")
-                    else:
-                        st.warning("Columns not found in the dataset or the question format is incorrect.")
-                elif 'column' in question.lower():
-                    # This is a request for a line graph (single column)
-                    column = question.split('of')[-1].strip()  # Handle 'of' keyword
-                    if column in df.columns:
-                        # Prepare the data for Plotly (line graph)
-                        column_data = df[column].dropna()  # Drop NaN values
-
-                        # Create the line plot
-                        fig = px.line(x=column_data.index, y=column_data, title=f"Graph of column '{column}'")
-                        st.plotly_chart(fig, use_container_width=True)
-                        st.success(f"Here is the graph of column '{column}'.")
+                    # Use TAPAS model to process the question
+                    result = tqa(table=df, query=question)
+
+                    # Display the raw output from TAPAS
+                    st.write("TAPAS Raw Output (Response):")
+                    st.write(result)  # This will display the raw output from TAPAS
+
+                    # Optionally, you can output the raw output as plain text:
+                    st.text("Raw TAPAS Output (Plain Text):")
+                    st.text(str(result))  # This will display raw output as plain text
+
+                    # Check if TAPAS is returning the expected answer
+                    answer = result.get('answer', None)
+                    if answer:
+                        st.write(f"TAPAS Answer: {answer}")
                     else:
-                        st.warning(f"Column '{column}' not found in the data.")
-                else:
-                    st.warning("Please ask a valid graph-related question (e.g., 'make a graph between column1 and column2').")
+                        st.warning("TAPAS did not return a valid answer.")
 
-        except Exception as e:
-            st.warning(f"Error processing question or generating graph: {str(e)}")
+            except Exception as e:
+                st.warning(f"Error processing question or generating answer: {str(e)}")