Update app.py

app.py

@@ -4,72 +4,77 @@ import seaborn as sns
 import matplotlib.pyplot as plt
 from dataset_import import load_data
 from data_cleaning import clean_data
-from data_labelling import label_dataset
+from YTCommentScraper import get_transcript, summarize_with_langchain_and_openai
 
 st.title("AnalyzeYT Dataset Analysis")
 
-# Load and clean dataset
-data = load_data()
+# Option to extract YouTube comments
+extract_comments = st.radio("Do you want to extract comments from a YouTube video?", ("No", "Yes"))
 
-if data is not None:
-    st.write("Loaded Data Preview:")
-    st.write(data.head())
-    
-    # Clean data
-    data = clean_data(data)
-    
-    st.write("Cleaned Data Preview:")
-    st.write(data.head())
-    
-    # Show data description
-    st.write("Data Description:")
-    st.write(data.describe())
+if extract_comments == "Yes":
+    openai_api_key = st.text_input("Enter your OpenAI API Key:", type='password')
+    video_id = st.text_input("Enter the YouTube Video ID:")
 
-    # Data visualization options
-    st.write("Data Visualization:")
-    chart_type = st.selectbox("Select Chart Type", ['Line Chart', 'Bar Chart', 'Histogram'])
+    if st.button('Extract and Summarize Comments'):
+        if openai_api_key and video_id:
+            try:
+                transcript, language_code = get_transcript(video_id)
+                summary = summarize_with_langchain_and_openai(transcript, language_code, openai_api_key)
+                st.write("Summary of YouTube Comments:")
+                st.markdown(summary)
+            except Exception as e:
+                st.write(str(e))
+        else:
+            st.write("Please enter both your OpenAI API Key and YouTube Video ID.")
+else:
+    # Proceed with normal data analysis
+    data = load_data()
 
-    if chart_type == 'Line Chart':
-        x_axis = st.selectbox("Select X-axis Column", data.columns)
-        y_axis = st.selectbox("Select Y-axis Column", data.columns)
-        title = st.text_input("Enter Chart Title", "Line Chart")
-        st.line_chart(data[[x_axis, y_axis]])
-        st.write(f"Line Chart: {title}")
-
-    elif chart_type == 'Bar Chart':
-        x_axis = st.selectbox("Select X-axis Column", data.columns)
-        y_axis = st.selectbox("Select Y-axis Column", data.columns)
-        title = st.text_input("Enter Chart Title", "Bar Chart")
-        st.bar_chart(data[[x_axis, y_axis]])
-        st.write(f"Bar Chart: {title}")
+    if data is not None:
+        st.write("Loaded Data Preview:")
+        st.write(data.head())
+        
+        # Clean data
+        data = clean_data(data)
+        
+        st.write("Cleaned Data Preview:")
+        st.write(data.head())
+        
+        # Show data description of cleaned data
+        st.write("Cleaned Data Description:")
+        st.write(data.describe())
 
-    elif chart_type == 'Histogram':
-        selected_column = st.selectbox("Select Column for Histogram", data.columns)
-        bins = st.slider("Number of Bins", min_value=10, max_value=100, value=30)
-        title = st.text_input("Enter Chart Title", "Histogram")
-        plt.hist(data[selected_column], bins=bins)
-        plt.title(title)
-        plt.xlabel(selected_column)
-        plt.ylabel('Frequency')
+        # Add correlation matrix for cleaned data
+        st.write("Correlation Matrix of Cleaned Data:")
+        corr = data.corr()
+        st.write(corr)
+        sns.heatmap(corr, annot=True, cmap='coolwarm')
         st.pyplot()
 
-    # Add option for labeling the dataset
-    st.write("Labeling Options:")
-    label_option = st.radio("Do you want to label your dataset?", ('No', 'Yes'))
+        # Data visualization options for cleaned data
+        st.write("Data Visualization on Cleaned Data:")
+        chart_type = st.selectbox("Select Chart Type", ['Line Chart', 'Bar Chart', 'Histogram'])
 
-    if label_option == 'Yes':
-        # Select columns for labeling
-        columns_to_label = st.multiselect("Select Columns to Label", data.columns)
-        
-        # Input new column names for labeled data
-        new_column_names = []
-        for col in columns_to_label:
-            new_name = st.text_input(f"Enter new name for labeled column '{col}':", f"{col}_labeled")
-            new_column_names.append(new_name)
-        
-        output_name = st.text_input("Enter Output File Name", "labeled_dataset.csv")
-        if st.button("Run Labeling"):
-            labeled_data = label_dataset(data, columns_to_label, new_column_names)
-            labeled_data.to_csv(output_name, index=False)
-            st.write("Labeling Completed. Download your file below:")
-            st.download_button(label="Download Labeled Dataset", data=labeled_data.to_csv(index=False), file_name=output_name, mime='text/csv')
+        if chart_type == 'Line Chart':
+            x_axis = st.selectbox("Select X-axis Column", data.columns)
+            y_axis = st.selectbox("Select Y-axis Column", data.columns)
+            title = st.text_input("Enter Chart Title", "Line Chart")
+            st.line_chart(data[[x_axis, y_axis]])
+            st.write(f"Line Chart: {title}")
+
+        elif chart_type == 'Bar Chart':
+            x_axis = st.selectbox("Select X-axis Column", data.columns)
+            y_axis = st.selectbox("Select Y-axis Column", data.columns)
+            title = st.text_input("Enter Chart Title", "Bar Chart")
+            st.bar_chart(data[[x_axis, y_axis]])
+            st.write(f"Bar Chart: {title}")
+
+        elif chart_type == 'Histogram':
+            selected_column = st.selectbox("Select Column for Histogram", data.columns)
+            bins = st.slider("Number of Bins", min_value=10, max_value=100, value=30)
+            title = st.text_input("Enter Chart Title", "Histogram")
+            plt.hist(data[selected_column], bins=bins)
+            plt.title(title)
+            plt.xlabel(selected_column)
+            plt.ylabel('Frequency')
+            st.pyplot()