Azure Notebooks is a Microsoft Azure Platform as a Service<\/u> (PaaS) offering of Jupyter Notebooks. A \u201cJupyter\u201d Notebook is an open-source web application that lets you create and share documents containing live code, equations, visualizations, and narrative text. You can use ADS notebooks for data cleaning and transformation, numerical simulation, statistical modeling, data visualization, and machine learning (just to name a few).<\/p>\n
Steps to create an ADS Notebook<\/strong><\/p>\n To create an ADS notebook (from within Azure Data Studio), you simply select File<\/strong>, then New Notebook<\/strong>. Once you create a notebook, you\u2019ll need to select a \u201cnotebook kernel\u201d for the notebook to \u201cuse\u201d.<\/p>\n What is a notebook kernel?<\/strong><\/p>\n A notebook kernel is\u00a0the \u201cengine\u201d that executes the code you add to the notebook. There are kernels for many languages available, for example the \u201cipython\u201d kernel (listed as \u201cPython 3\u201d) executes Python code. By default, the SQL kernel (for executing T-SQL queries for SQL Server) is selected.<\/p>\n In this example, we\u2019ll change the kernel to Python 3. Once you select Python 3, you will need to configure Python for the notebook, by selecting the option to use either an existing<\/em> Python installation (if Python is already installed) or create a new Python installation (if you select this option, Azure Data Studio will perform the Python installation for you). Once you select a kernel and save the notebook, when you reopen the notebook, the associated kernel will be automatically launched and be ready for use.<\/p>\n Adding Cells<\/strong><\/p>\n Notebooks consist of a series of \u201ccells\u201d. Cells can be either TEXT or CODE. A cell is essentially a section or \u201cstep\u201d in a notebook. You can interleave code and text, but they are separate elements that format in whatever order you’ve placed them. Text cells allow you to easily document code in a notebook by adding \u201cMarkdown text blocks\u201d in between code cells. Like GitHub, Azure Data Studio notebooks use markdown language for formatting text cells.<\/p>\n Let\u2019s start building this notebook by adding a new text cell by clicking the\u00a0+Cell<\/strong>\u00a0command in the toolbar and selecting\u00a0Text cell<\/strong>. The cell opens up in \u201cedit mode\u201d so you can type some (markdown) text. You can select either a rich text, markdown or \u201csplit view\u201d (of the cell) to see as you type text, which helps you to get the formatting just right.<\/p>\n Linking to External Information<\/strong><\/p>\n One of the \u201ccooler things\u201d about notebook text cells is that you can include a reference link <\/em>within your text. So, for example, in the above image you can see that I\u2019ve added a text cell with a heading and then underneath, an additional note about Azure Notebooks. Since my audience may not have used notebooks before, I have inserted a reference link to a web page where they can read more information about Azure notebooks. To add a link, you just highlight the desired text and click the link icon, where you can paste in the appropriate URL:<\/p>\n Code Cells<\/strong><\/p>\n Notebook code cells allow you to enter and run program code interactively within a notebook. You add a new code cell by clicking the\u00a0+Cell<\/strong>\u00a0command in the toolbar and then selecting\u00a0Code cell<\/strong>. Below, I\u2019ve added a code cell to import the python modules that I want to use later in the notebook (note that I have already installed these libraries into my environment).<\/p>\n If you are unsure if a module is installed, you can click on Manage packages <\/strong>(shown below) and then view all of the installed packages or, install something new by click Add new<\/strong>:<\/p>\n Getting Data<\/strong><\/p>\n pyodbc is an open-source Python module that makes accessing ODBC (Open Database Connectivity) databases simple. Using pyodbc, you can easily connect Python applications to data sources with an ODBC driver.\u00a0In the code cell below, since I know it\u2019s already installed and we have imported it in the cell above, \u00a0I\u2019ve created both a connection to a SQL Server database named \u201cJimsData\u201d and a SQL command string to query a table named Correlations.<\/p>\n In the code cell below, I execute the SQL command,\u00a0 return the data into a pandas DataFrame object (named \u201cmydf\u201d) and, just for clarity, the DataFrame is then printed using the command print(mydf):<\/p>\n More Commentary<\/strong><\/p>\n Even though I\u2019ve added markdown text between each of the \u201csteps\u201d in my Azure notebook, notice that I always include a number of comments within my code cells, typically explaining each logical or functional purpose. Between the markdown and the code comments, hopefully everything is quite clear.<\/p>\n Running Some Code<\/strong><\/p>\n To run a single code cell, you click\u00a0Run cell<\/strong>\u00a0(the round black arrow) to the left of the cell or select the cell and press F5. You can run all the cells in the notebook by clicking\u00a0Run all<\/strong>\u00a0in the toolbar. If you \u201crun all\u201d the cells are run one at a time and execution will stop if an error is encountered in a cell. The results (or outputs) from a code cell are shown below the cell. You can \u201cclear\u201d the results of all the executed cells in the notebook by selecting the\u00a0Clear Results<\/strong>\u00a0button in the toolbar. The image below shows the above code cell and its output (the printed DataFrame) below it.<\/p>\n In the next step, I added another text cell with a link to documentation on the python \u201ccorr\u201d function, which is then used in the code cell beneath it to create a correlation matrix – using the data sourced from SQL Server \u2013 which we previously loaded into a DataFrame. A correlation matrix is\u00a0a table showing correlation coefficients between variables<\/strong>. Each value in the table shows the correlation between two variables. A correlation matrix is used to summarize data, as an input into a more advanced analysis, and as a diagnostic for advanced analyses.<\/p>\n Below is the text cell and then the code cell followed by its output.<\/p>\n Visualizing the Correlation Matrix<\/strong><\/p>\n A heatmap\u00a0contains values representing various shades of the same color for each value to be plotted. Usually, the darker shades of the chart represent higher values than the lighter shade. A heatmap is often a good choice to better illustrate the correlation between variables based upon a correlation matrix. Like the previous group of cells, below I\u2019ve added a text cell referencing the seaborn heatmap function, followed by a code cell that implements that function using our generated correlation matrix (from the previous cell). The last line prints the heatmap.<\/p>\n Again, we can click the cell run button to see the output of this code cell which is a generated heatmap of variable correlations:<\/p>\n As a last step, I added a final text cell with a bit of a \u201cconclusion\u201d:<\/p>\n Final Thoughts<\/strong><\/p>\n Azure Notebooks are a great option for sharing solutions to challenging problems since both working code and notes can easily be shared between individuals or teams for review, comment, and revision. Another valuable use case is to use Azure Notebooks to document a process or create a how-to guide for an important procedure.<\/p>\n","protected":false},"excerpt":{"rendered":" The objective of this exercise is to demonstrate creating an Azure Data Studio (ADS) Notebook using Python to query SQL Server and investigate correlations between variables. Variable correlation can be key to increasing the accuracy of a forecast or predicting potential outcome scenarios. What is an Azure Data Studio Notebook? Azure Notebooks is a Microsoft…<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[27],"tags":[],"class_list":["post-3586","post","type-post","status-publish","format-standard","hentry","category-quebit-value"],"acf":[],"yoast_head":"\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ1-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ2-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ3-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ4-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ5-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ6-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ7-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ8-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ9-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ10-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ11-min.jpg\")
<\/p>\n![\"\"](\"https:\/\/quebit.com\/askquebit\/wp-content\/uploads\/2021\/12\/AZ12-min.jpg\")
<\/p>\n