Write and run SQL and Python code in a Snowflake Notebook¶

Import packages to use in notebooks¶

Snowflake Notebooks include Streamlit and the third-party packages listed in the Snowflake Anaconda channel.

After your organization administrator accepts the terms, you can import libraries to use in Snowflake Notebooks:

1. Sign in to Snowsight.

2. Select Notebooks.

3. Select a specific notebook for which you want to install Python packages.

4. (Optional) Select Packages.

5. Search for packages listed in the Snowflake Anaconda channel, such as numpy, pandas, requests, and urllib3.

6. Select a package to install it for use in your notebook, and optionally change the default package version in the list of Installed Packages.

   Packages installed by you appear under Installed Packages.

7. To use the previously installed libraries, add import statements to the code in a Python cell.

   For example, to use the previously installed scikit-learn package, add an import statement for that package at the beginning of your code:

   import scikit-learn
   

   Copy

Visualize results by using Streamlit¶

You can use any chart elements supported by Streamlit version 1.22.0 to create a line chart, bar chart, area chart, scatter plot, or a map with points on it. See Chart elements in the API reference of the Streamlit library documentation.

For example, to reference a SQL cell CELL1 that outputs two columns of data as a table, where CELL1 includes the following SQL code:

SELECT 'FRIDAY' as SNOWDAY, 0.2 as CHANCE_OF_SNOW
UNION ALL
SELECT 'SATURDAY',0.5
UNION ALL
SELECT 'SUNDAY', 0.9;

To visualize the results in a line chart, you can use the following Python code to reference the results of the SQL cell, convert it to a Snowpark DataFrame and then a Pandas DataFrame, then display a line chart:

import streamlit as st

my_df = cells.CELL1.to_df().to_pandas()

st.line_chart(my_df, x='SNOWDAY', y='CHANCE_OF_SNOW')

After you run both cells, the following visualization appears:

For details on referencing other cells, see Reference results and chain cells in a notebook.

Visualize results by using Altair¶

To visualize results using Altair visualizations, use the Streamlit st.altair_chart element.

• For details on the Altair chart element, see st.altair_chart.

• For details on available visualization types when using Altair, see Vega-Altair: Declarative Visualization in Python.

Snowflake Notebooks currently support Altair version 4.0.

For example, to display a heat map based on the output of SQL cell CELL3, where CELL3 includes the following SQL code:

SELECT 1 as "a", 1 as "b", 35 as "c"
UNION ALL
SELECT 17, 7, 2
UNION ALL
SELECT 14, 29, 18;

Write the following Python code in a Python cell to reference the output of the SQL cell, process the output into a Snowpark DataFrame and then a Pandas DataFrame, create an Altair chart object, and then use Streamlit to display the chart:

import streamlit as st
import altair as alt

my_df = cells.CELL3.to_df().to_pandas()

chart = (
  alt.Chart(my_df)
  .mark_rect()
  .encode(x="a", y="b", color="c")
)

st.altair_chart(chart, use_container_width=False, theme=None)

After you run both cells, the following visualization appears:

For details on referencing other cells, see Reference results and chain cells in a notebook.

Visualize results with Matplotlib¶

To visualize results using Matplotlib visualizations, use the Streamlit st.pyplot element.

To use a Matplotlib visualization, install the Matplotlib library for your notebook:

1. From the notebook, select Packages.

2. Locate the matplotlib library and select the library to install it.

3. Write code to visualize your results.

For example, to visualize a bar chart of a random array of numbers, add the following code to a Python cell:

import streamlit as st
import matplotlib.pyplot as plt
import numpy as np

arr = np.random.normal(1, 1, size=100)
fig, ax = plt.subplots()
ax.hist(arr, bins=20)

st.pyplot(fig)

After you run the cell, the following visualization appears:

For more details on using the st.pyplot chart element, see st.pyplot.

Visualize results with seaborn¶

To visualize results using seaborn visualizations, use the Streamlit st.pyplot element in combination with the seaborn library.

To use a seaborn visualization, you must install both the seaborn library and the Matplotlib library for your notebook:

1. From the notebook, select Packages.

2. Locate the seaborn library and select the library to install it.

3. Locate the matplotlib library and select the library to install it.

4. Write code to visualize your results.

To use seaborn, use the st.pyplot, but instead of passing a Matplotlib figure into the fig parameter, pass a seaborn figure.

For example, to display a bar chart based on the output of SQL cell CELL0, where CELL0 includes the following SQL code:

SELECT 1 as "a", 1 as "b", 35 as "c"
UNION ALL
SELECT 17, 7, 2
UNION ALL
SELECT 14, 29, 18;

Write the following Python code in a Python cell:

import streamlit as st
import seaborn as sns
import matplotlib.pyplot as plt

f, ax = plt.subplots(figsize=(3, 3))
fig=f.figure

my_df = cells.CELL0.to_df().to_pandas()

sns.set_color_codes("pastel")
sns.barplot(x="a", y="b", data=my_df,
            label="Total", color="c")

st.pyplot(fig)

After you run both cells, the following visualization appears:

For more details on using the st.pyplot chart element, see st.pyplot.

For examples of seaborn visualizations, see the seaborn Example gallery.

Warehouse recommendations for running Snowflake Notebooks¶

When you create a notebook, you select a warehouse in which to run the Snowflake Notebook processes and any code. Snowflake recommends using an X-Small warehouse to minimize credit consumption.

You might want to run large queries using SQL or perform compute-intensive operations using Snowpark Python that require a larger warehouse. In this case, you have two options:

• Specify a warehouse to use for a specific cell or set of cells.

  This ensures that the queries in those cells use the larger warehouse, but notebook processes continue to run on the smaller notebook warehouse, so you’ll use less compute when you’re not running queries.

  For example, to specify a warehouse called notebooks to use for specific cells, you can add a SQL cell with the USE WAREHOUSE command or add the command to the Python code in a Python cell:

  SQL::

  USE WAREHOUSE notebooks;
  

  Copy

  Python::

  session.sql('USE WAREHOUSE notebooks;').collect()
  

  Copy

  After you run any cell specifying USE WAREHOUSE, all cells that run in the remainder of the notebook session use the specified warehouse. If you no longer need queries to run with a large warehouse, write a new cell to specify a smaller warehouse to use.

• Choose a larger warehouse for the entire notebook.

  If you choose this option, all queries that run in the notebook have more compute available, but the warehouse runs for the entire notebook session, even if queries do not run.

  조심

  Choosing this option can dramatically increase credit consumption for the entire notebook.

  To change the notebook warehouse:

  1. Select the notebook title, and then select Notebook settings.

  2. Select a new warehouse to use.