A desktop data table holds rows with named columns. Create new data table constructs a table, Merge data tables appends rows, and Filter data table retains the rows satisfying a condition. Appending does not automatically reconcile duplicate business keys. Two identical rows and two conflicting rows with the same ID need different policies.
These exercises separate merge from deduplication. Delete duplicate rows from data table removes rows whose values and column types match. It does not choose the newest status among conflicting versions of an ID. State the row grain before choosing the transformation. Preserve known zero quantities and text identifiers. The checked worksheet effects include the full output rows, so a plausible row count cannot hide a lost column or a wrongly retained record. Local tables are independent fictional fixtures rather than live workbook connections.