Upload 2 files

data_science_with_promt.py

@@ -0,0 +1,111 @@
+import datahorse
+import streamlit as st
+import pandas as pd
+import matplotlib.pyplot as plt
+import io
+import sys
+
+st.title('Text to Data Analysis')
+
+# Initialize session state for conversation history
+if 'conversation' not in st.session_state:
+    st.session_state.conversation = []
+
+uploaded_file = st.file_uploader('Upload .csv data: ')
+
+
+query = st.text_input('Search about data: ')
+
+def capture_output(func):
+    # Capture stdout
+    old_stdout = sys.stdout
+    new_stdout = io.StringIO()
+    sys.stdout = new_stdout
+    
+    # Capture matplotlib figures
+    fig = plt.figure()
+    
+    result = func()
+    
+    # Reset stdout
+    sys.stdout = old_stdout
+    output = new_stdout.getvalue()
+    
+    # Check if a plot was created
+    if plt.gcf().axes:
+        return fig
+    elif output:
+        return output
+    else:
+        return result
+
+if uploaded_file is not None:
+
+    # Add some example queries
+    st.sidebar.header("Example Queries")
+    st.sidebar.write("1. Show me a summary of the data")
+    st.sidebar.write("2. Create a bar chart of [column_name]")
+    st.sidebar.write("3. What is the average of [column_name]?")
+    st.sidebar.write("4. Show me the correlation between [column1] and [column2]")
+    st.sidebar.write("5. List the unique values in [column_name]")
+    st.sidebar.write("6. What is the maximum value in [column_name]?")
+    st.sidebar.write("7. Create a line chart of [column_name] over time")
+    st.sidebar.write("8. How many missing values are in [column_name]?")
+    st.sidebar.write("9. Filter rows where [column_name] is greater than [value]")
+    st.sidebar.write("10. Generate a pie chart for [column_name]")
+
+
+    df = datahorse.read(uploaded_file)
+
+    col1, col2 = st.columns(2)
+
+
+    if col1.button('Search it'):
+        # Append user query to conversation history
+        user_message = f"""
+        <div style='background-color: #f0f0f0; padding: 10px; border-radius: 10px; margin: 5px 0;'>
+            <strong>Me:</strong> {query}
+        </div>
+        """
+        st.session_state.conversation.append(user_message)
+        
+        # Get response from datahorse
+        response = capture_output(lambda: df.chat(query))
+        
+        st.subheader("Response:")
+        
+        if isinstance(response, plt.Figure):
+            st.pyplot(response)
+        elif isinstance(response, str):
+            st.text(response)
+        elif isinstance(response, pd.DataFrame):
+            st.dataframe(response)
+        elif response is not None:
+            st.write(response)
+        else:
+            st.write("No output was captured.")
+
+        # Append response to conversation history
+        response_message = f"""
+        <div style='background-color: #4CAF50; color: white; padding: 10px; border-radius: 10px; margin: 5px 0;'>
+            <strong>Datahorse:</strong> Response displayed above
+        </div>
+        """
+        st.session_state.conversation.append(response_message)
+    # Clear conversation history
+    if col2.button('Clear Conversation'):
+        st.session_state.conversation = []
+
+
+# Display conversation history
+for message in reversed(st.session_state.conversation):
+    st.markdown(message, unsafe_allow_html=True)
+
+# Display the dataframe
+if uploaded_file is not None:
+    st.write("Preview of the uploaded data:")
+    st.dataframe(df.head())
+    
+    # Display column names
+    st.write("Column names in the dataset:")
+    st.write(", ".join(df.columns))

requirements.txt

@@ -0,0 +1,6 @@
+streamlit
+datahorse
+pandas
+matplotlib
+io
+sys