Requirements:
Parallel processing # PySpark runs distributed across Spark clusters.
Minimize duplication of data # OneLake shortcut + Spark queries avoid unnecessary copies.
Time series analysis, anomaly detection, visualization # can be done in PySpark within Fabric Notebooks.
pandas (C) is single-threaded and not efficient for billions of items.
Power BI visuals (A) are good for visualization but not for transformation or large-scale anomaly detection.
Correct: B.
Reference: Use PySpark in Fabric notebooks

Tabular Editor is an advanced tool for editing Tabular models outside of Power BI Desktop that allows you to script out changes and apply them across multiple columns or tables. To accomplish the task programmatically, you would:
* Open the model in Tabular Editor.
* Create an Advanced Script using C# to iterate over all tables and their respective columns.
* Within the script, check if the column name ends with 'Key'.
* For columns that meet the condition, set the properties accordingly: IsHidden = true, IsNullable = false, SummarizeBy = None, IsAvailableInMDX = false.
* Additionally, mark the column as a key column.
* Save the changes and deploy them back to the Fabric tenant.
References: The ability to batch-edit properties using scripts in Tabular Editor is well-documented in the tool's official documentation and user community resources.

To ensure all rows appear in the Python visual within a Power BI report, option C, adding a unique field to each row, is the correct solution. This will prevent automatic grouping by unique values and allow for all instances of data to be represented in the visual. References = For more on Power BI Python visuals and how they handle data, please refer to the Power BI documentation.

The pickupLongitude column has duplicate values. This can be inferred because the 'Distinct count' is 935 while the 'Count' is 1000, indicating that there are repeated values within the column. References = Microsoft Power BI documentation on data profiling could provide further insights into understanding and interpreting column statistics like these.

Explanation:

Comprehensive Detailed Explanation
We need to write a query that summarizes order quantities by year and product, and also include the yearly total for all products in each row.
Step 1: What the query needs
Extract the year from SalesOrderDetail.ModifiedDate.
Join with the Product table to get the product name.
Aggregate with SUM(OrderQty).
Return grouped data by year and product, with an additional subtotal row per year (all products combined).
Step 2: Evaluate the SELECT Clause
We must extract the year using:
YEAR(so.ModifiedDate) AS OrderDate
This converts the ModifiedDate column into a year value for grouping.
Step 3: Evaluate the GROUP BY options
CUBE(YEAR, P.Name) # Produces all combinations of year totals, product totals, and grand totals. Too many combinations, not required.
GROUPING SETS # Could achieve the result but requires explicitly listing the sets. Less direct.
ROLLUP(YEAR, P.Name) # Produces grouping by (Year, Product) and then a subtotal per Year. Exactly what is required.
YEAR only # Would group only by year, losing per-product breakdown.
Correct: ROLLUP(YEAR(so.ModifiedDate), P.Name)
Step 4: Completed Query
SELECT
YEAR(so.ModifiedDate) AS OrderDate,
p.Name,
SUM(so.OrderQty) AS OrderQty
FROM dbo.SalesOrderDetail so
INNER JOIN dbo.Product p
ON p.ProductID = so.ProductID
GROUP BY ROLLUP(YEAR(so.ModifiedDate), p.Name);
Why This Works
YEAR(so.ModifiedDate) extracts year for grouping.
ROLLUP(YEAR, P.Name) provides both product-level totals and yearly subtotals.
Ensures the requirement: "include the yearly sum of order quantities for all the products in each row." References GROUP BY ROLLUP in T-SQL Aggregate functions in Microsoft Fabric warehouses