Update app.py

app.py

@@ -3,6 +3,22 @@ import streamlit as st
 from transformers import TapasForQuestionAnswering, TapasTokenizer, T5ForConditionalGeneration, T5Tokenizer
 import torch
 
+# Assuming df is uploaded or pre-defined (you can replace with actual data loading logic)
+# Example DataFrame (replace with your actual file or data)
+data = {
+    'Column1': [1, 2, 3, 4],
+    'Column2': [5.5, 6.5, 7.5, 8.5],
+    'Column3': ['a', 'b', 'c', 'd']
+}
+df = pd.DataFrame(data)
+
+# Check if DataFrame is valid
+if df is not None and not df.empty:
+    # Select numeric columns
+    df_numeric = df.select_dtypes(include='number')
+else:
+    df_numeric = pd.DataFrame()  # Empty DataFrame if input is invalid
+
 # Load TAPAS model and tokenizer
 tqa_model = TapasForQuestionAnswering.from_pretrained("google/tapas-large-finetuned-wtq")
 tqa_tokenizer = TapasTokenizer.from_pretrained("google/tapas-large-finetuned-wtq")
@@ -11,9 +27,6 @@ tqa_tokenizer = TapasTokenizer.from_pretrained("google/tapas-large-finetuned-wtq
 t5_model = T5ForConditionalGeneration.from_pretrained("t5-small")
 t5_tokenizer = T5Tokenizer.from_pretrained("t5-small")
 
-# Assuming 'df' is the DataFrame you are using and has numeric columns
-df_numeric = df.select_dtypes(include='number')
-
 # User input for the question
 question = st.text_input('Type your question')
 
@@ -60,6 +73,10 @@ with st.spinner():
             st.warning("Please retype your question and make sure to use the column name and cell value correctly.")
 
 
+# Assuming 'column_name' exists and is selected or provided by the user
+# Example of getting 'column_name' from user input (adjust this part according to your app):
+column_name = st.selectbox("Select a column", df.columns)
+
 # Manually fix the aggregator if it returns an incorrect one
 if 'MEDIAN' in question.upper() and 'AVERAGE' in aggregator.upper():
     aggregator = 'MEDIAN'