Data Transformation
Reshaping and summarizing
Transformation tools change the shape or the row count of your data. Summarize groups by chosen fields and applies actions such as Sum, Count, Count Non Null, Min, Max, First, Last, and Concat (set the separator explicitly). Cross Tab pivots rows to columns: pick Group By fields, a Header field whose values become new field names, a Data field, and an aggregation method. Transpose does the opposite, turning the fields listed as Data Fields into rows of Name/Value pairs while Key Fields are carried through. Cross Tab and Transpose are how you move between wide and long layouts.
Row-level tools do the rest. Filter splits into True/False outputs; Sort orders by multiple fields; Sample takes First N, Last N, Random N, Skip N, or 1 of every N; Unique keeps the first record per combination of unique fields and sends the rest to the Duplicate output; Record ID numbers rows; Multi-Row Formula references other rows (negative offsets look backward) with a Group By and a setting for rows that do not exist; Running Total and Tile handle cumulative and bucketing logic.
Traps
Cross Tab header values become field names, so messy values create messy fields, and missing group/header combinations produce nulls. Summarize output field types may differ from the input. Unique keeps the first record as it arrives, so sort first if order matters.
Rule of thumb
Wide-to-long is Transpose; long-to-wide is Cross Tab. Summarize reduces rows; never use it expecting the original row count.