Defining SciPyR Models

SciPYR

Infoveave supports custom analysis with Python on the SciPyR workbench to run large statistical models. The SciPyR workbench helps to define accurate custom analysis easily using existing Infoveave datasources.

Accessing SciPyR Book

To perform SciPyR analysis in Infoveave, follow these steps

1. Access the SciPyR section by clicking on Analysis SciPyR. You will see a list of all SciPyR books, including those created by you or shared with you.
2. To create a new SciPyR book based on Python, click on New SciPyR Book button . New SciPyR Book dialog box will display.

Selecting Python workbook

Note

• You can select the checkbox option on Optional Upload of Jupyter Book to upload a Jupyter notebook file directly.
• For Jupyter Notebook uploads, it’s essential to note that Infoveave only accepts .ipynb files.
• When using the Jupyter Notebook the workspace will be loaded with the content.

Initiate the creation of a new Python Workbook by providing a unique name.

1. Assign a Name for SciPyR Workbook.
2. Click on Save to start creating your custom analysis in Infoveave. This will directly take you to the SciPyR workbook.
3. Upon opening the SciPyR Workbook, you can begin writing your Python program directly in the Workbook.
4. To start writing your Python program, select the default cell and begin writing.
5. To expand your analysis, add new cells by simply clicking the plus icon + near the drop-down.
6. Use the drop-down menu in the Workbook to select Code when writing Python code and Markdown when adding headings or titles.

Note

• When you add a title using Markdown in the SciPyR Workbook, it will be automatically saved as a summary under the Summary Panel.
• During Python coding, any comments you include in the code will also be automatically added as a summary in the Summary Panel.

Connecting to datasource for exploratory data analysis

Integrate SQL queries to your analysis through the Insert Query feature.

To insert SQL queries on DataSources into the SciPyR Workbook, follow the below steps

1. Click on the Insert Query icon in the SciPyR Workbook. Insert Query popup comes up.
2. From the dropdown menu, select the desired data source that contains the SQL data.
3. Write the SQL query you need in the provided workspace area.
4. Click on the Execute query icon to execute the SQL query and see the results.
5. If you want to insert the query to the SciPyR Workbook, click the Save button after executing the query.
6. After this run the query. This will insert the query and its results into the Workbook for further analysis and reporting.

Note

When you utilize the Insert Query feature to add a query into the SciPyR Workbook, it will be automatically inserted in the following syntax

df = infoveave.query("34e44add-3bae-4ae6-a096-9888c52ada29")

• df This is the variable name chosen to store the data retrieved from the query. It stands for DataFrame, which is a data structure used in Python to store tabular data.
• "34e44add-3bae-4ae6-a096-9888c52ada29" This is the unique address or identifier of the specific query within Infoveave. It acts as a reference to locate and fetch the desired data from the dataset associated with that query.

Tip

The Date option in the Insert Query dialog box will appear based on the structure of your query. If your query includes date-related data, you’ll be able to select the “Date” option to handle date values appropriately within the SciPyR Workbook.

Editing the inserted query

Click on Edit Query icon .

Note

The Edit query option is functional only when a query has been inserted into the specific cell.

Building model & generating insights

Customize your analysis by adding, editing, and running cells, focusing on building your machine learning model within the SciPyR Workbook. Utilize the flexibility of manipulating cells to refine and tailor your analysis, leading to the generation of valuable insights from the data.

• Customize the analysis by adding and editing cells within the SciPyR Workbook.
• Incorporate Python code specific to your analysis requirement.
• Execute individual cells, by simply selecting the cell and clicking on the Run icon .
• You can also run the entire set of cells together by clicking the icon, to execute the complete analysis.
• To refresh the kernel, click on the icon.
• To save the SciPyR as a PDF file in the local storage, click on View PDF button.
• To save the analysis, click Save.
• Go to Analysis SciPyR to view the saved analysis.

You have the flexibility to add cells above or below existing ones, move cells around, copy/duplicate cells, and delete cells to customize your analysis as needed.

Note

You can view the tables generated (if any) from variables section.