Update app.py

app.py

@@ -77,22 +77,42 @@ else:
                 coordinates = raw_answer.get('coordinates', [])
                 cells = raw_answer.get('cells', [])
                 
+                # Check if the answer contains non-numeric values, and filter them out
+                numeric_cells = []
+                for cell in cells:
+                    try:
+                        numeric_cells.append(float(cell))  # Convert to float if possible
+                    except ValueError:
+                        pass  # Ignore non-numeric cells
+
                 # Handle aggregation based on user question or TAPAS output
                 if 'average' in question.lower() or aggregator == 'AVG':
-                    avg_value = df_numeric.mean().mean()  # Calculate average of the entire DataFrame
-                    base_sentence = f"The average for '{question}' is {avg_value:.2f}."
+                    if numeric_cells:
+                        avg_value = sum(numeric_cells) / len(numeric_cells)  # Calculate average
+                        base_sentence = f"The average for '{question}' is {avg_value:.2f}."
+                    else:
+                        base_sentence = f"No numeric data found for calculating the average of '{question}'."
                 elif 'sum' in question.lower() or aggregator == 'SUM':
-                    total_sum = df_numeric.sum().sum()  # Calculate sum of all numeric values
-                    base_sentence = f"The sum for '{question}' is {total_sum:.2f}."
+                    if numeric_cells:
+                        total_sum = sum(numeric_cells)  # Calculate sum
+                        base_sentence = f"The sum for '{question}' is {total_sum:.2f}."
+                    else:
+                        base_sentence = f"No numeric data found for calculating the sum of '{question}'."
                 elif 'max' in question.lower() or aggregator == 'MAX':
-                    max_value = df_numeric.max().max()  # Calculate max value across DataFrame
-                    base_sentence = f"The maximum value for '{question}' is {max_value:.2f}."
+                    if numeric_cells:
+                        max_value = max(numeric_cells)  # Find max value
+                        base_sentence = f"The maximum value for '{question}' is {max_value:.2f}."
+                    else:
+                        base_sentence = f"No numeric data found for finding the maximum value of '{question}'."
                 elif 'min' in question.lower() or aggregator == 'MIN':
-                    min_value = df_numeric.min().min()  # Calculate min value across DataFrame
-                    base_sentence = f"The minimum value for '{question}' is {min_value:.2f}."
+                    if numeric_cells:
+                        min_value = min(numeric_cells)  # Find min value
+                        base_sentence = f"The minimum value for '{question}' is {min_value:.2f}."
+                    else:
+                        base_sentence = f"No numeric data found for finding the minimum value of '{question}'."
                 elif 'count' in question.lower() or aggregator == 'COUNT':
-                    row_count = df_numeric.count().sum()  # Count of all non-null values
-                    base_sentence = f"The total count of non-null values for '{question}' is {row_count}."
+                    count_value = len(numeric_cells)  # Count numeric cells
+                    base_sentence = f"The total count of numeric values for '{question}' is {count_value}."
                 else:
                     # Construct a base sentence for other aggregators or no aggregation
                     base_sentence = f"The answer from TAPAS for '{question}' is {answer}."