Skip to main content
Last updated on

Tutorial: Develop a Databricks app with Streamlit

This tutorial shows how to build a Databricks app using the Databricks SQL Connector for Python and Streamlit. You'll learn how to develop an app that does the following:

  • Reads a Unity Catalog table and displays it in a Streamlit interface.
  • Edits data and writes it back to the table.

Step 1: Configure privileges​

These examples assume that your app uses app authorization. Your app’s service principal must have:

  • SELECT privilege on the Unity Catalog table
  • MODIFY privilege on the Unity Catalog table
  • CAN USE privilege on the SQL warehouse

For more information, see Unity Catalog privileges reference and SQL warehouse ACLs.

Step 2: Install dependencies​

Create a requirements.txt file and include the following packages:

Text
databricks-sdk
databricks-sql-connector
streamlit
pandas

Step 3: Configure app execution​

Create an app.yaml file to define how your app starts in Databricks Apps.

YAML
command: ['streamlit', 'run', 'app.py']

Step 4: Read a Unity Catalog table​

This example code demonstrates how to read data from a Unity Catalog table and display it using Streamlit. Create an app.py file that fulfills the following objectives:

  • Uses app service principal authentication.
  • Prompts the user for the SQL warehouse HTTP path and Unity Catalog table name.
  • Executes a SELECT * query on the specified table.
  • Displays the result in a Streamlit st.dataframe.

app.py​

Python
import pandas as pd
import streamlit as st
from databricks import sql
from databricks.sdk.core import Config
import os

cfg = Config()

# Use app service principal authentication
def get_connection(http_path):
    server_hostname = cfg.host
    if server_hostname.startswith('https://'):
        server_hostname = server_hostname.replace('https://', '')
    elif server_hostname.startswith('http://'):
        server_hostname = server_hostname.replace('http://', '')
    return sql.connect(
        server_hostname=server_hostname,
        http_path=http_path,
        credentials_provider=lambda: cfg.authenticate,
        _use_arrow_native_complex_types=False,
    )

# Read data from a Unity Catalog table and return it as a pandas DataFrame
def read_table(table_name: str, conn) -> pd.DataFrame:
    with conn.cursor() as cursor:
        cursor.execute(f"SELECT * FROM {table_name}")
        return cursor.fetchall_arrow().to_pandas()

# Use Streamlit input fields to accept user input
http_path_input = st.text_input(
    "Enter your Databricks HTTP Path:", placeholder="/sql/1.0/warehouses/xxxxxx"
)
table_name = st.text_input(
    "Specify a Unity Catalog table name:", placeholder="catalog.schema.table"
)

# Display the result in a Streamlit DataFrame
if http_path_input and table_name:
    conn = get_connection(http_path_input)
    df = read_table(table_name, conn)
    st.dataframe(df)
else:
    st.warning("Provide both the warehouse path and a table name to load data.")

Step 5: Edit a Unity Catalog table​

This example code allows users read, edit, and write changes to a Unity Catalog table using Streamlit's data editing features. Add the following functionality to the app.py file:

  • Use INSERT OVERWRITE to write the updated data back to the table.

app.py​

Python
import pandas as pd
import streamlit as st
from databricks import sql
from databricks.sdk.core import Config
import math

cfg = Config()

# Use app service principal authentication
def get_connection(http_path):
    server_hostname = cfg.host
    if server_hostname.startswith('https://'):
        server_hostname = server_hostname.replace('https://', '')
    elif server_hostname.startswith('http://'):
        server_hostname = server_hostname.replace('http://', '')
    return sql.connect(
        server_hostname=server_hostname,
        http_path=http_path,
        credentials_provider=lambda: cfg.authenticate,
        _use_arrow_native_complex_types=False,
    )

# Read data from a Unity Catalog table and return it as a pandas DataFrame
def read_table(table_name: str, conn) -> pd.DataFrame:
    with conn.cursor() as cursor:
        cursor.execute(f"SELECT * FROM {table_name}")
        return cursor.fetchall_arrow().to_pandas()

# Format values for SQL, handling NaN/None as NULL
def format_value(val):
    if val is None or (isinstance(val, float) and math.isnan(val)):
        return 'NULL'
    else:
        return repr(val)

# Use `INSERT OVERWRITE` to update existing rows and insert new ones
def insert_overwrite_table(table_name: str, df: pd.DataFrame, conn):
    progress = st.empty()
    with conn.cursor() as cursor:
        rows = list(df.itertuples(index=False))
        values = ",".join([f"({','.join(map(format_value, row))})" for row in rows])
        with progress:
            st.info("Calling Databricks SQL...")
        cursor.execute(f"INSERT OVERWRITE {table_name} VALUES {values}")
    progress.empty()
    st.success("Changes saved")

# Use Streamlit input fields to accept user input
http_path_input = st.text_input(
    "Enter your Databricks HTTP Path:", placeholder="/sql/1.0/warehouses/xxxxxx"
)
table_name = st.text_input(
    "Specify a Unity Catalog table name:", placeholder="catalog.schema.table"
)

# Display the result in a Streamlit DataFrame
if http_path_input and table_name:
    conn = get_connection(http_path_input)
    if conn:
        st.success("✅ Connected successfully!")
        original_df = read_table(table_name, conn)
        edited_df = st.data_editor(original_df, num_rows="dynamic", hide_index=True)
        df_diff = pd.concat([original_df, edited_df]).drop_duplicates(keep=False)
        if not df_diff.empty:
            st.warning(f"⚠️ You have {len(df_diff) // 2} unsaved changes")
            if st.button("Save changes"):
                insert_overwrite_table(table_name, edited_df, conn)
                st.rerun()
else:
    st.warning("Provide both the warehouse path and a table name to load data.")

Next steps​

On this page