This QuickStart shows how to use Sigma's Transpose table to reshape data in two directions — converting rows into columns for side-by-side comparisons, and converting wide columns into rows for long-format analysis.

Most data arrives from the warehouse in one of two shapes that aren't analysis-ready:

Transpose addresses both. It operates directly on the data from your data platform and produces a reshaped output that you can then group, aggregate, filter, and chart within Sigma.

This QuickStart covers both directions using two Sigma sample datasets — PLUGS_ELECTRONICS for Row to Column, and FLIGHTS for Column to Row.

Along the way you'll learn how to:

For more information on Sigma's product release strategy, see Sigma product releases

If something doesn't work as expected, here's how to contact Sigma support

Target Audience

Analysts who need to reshape source data for comparison views, long-format analysis, or charting. Familiarity with Sigma workbooks and basic table operations is assumed.

Prerequisites

Sigma Free Trial

Footer

The Row to Column direction converts dimension values that are stacked in rows into separate columns — one column per unique value in the label column. Use this when you want a wide layout for side-by-side comparison.

The data:
PLUGS_ELECTRONICS_HANDS_ON_LAB_DATA from the RETAIL schema — a retail transaction dataset including Product Type, Cost, Store Region, and Store State.

The problem:
 Each transaction has a product category in a single Product Type column. To compare cost across categories by store location, each category needs to be its own column.

Set up the workbook

In Sigma, click Create New and select Workbook. Save the workbook as Transpose Tables QuickStart.

Rename the default page to Data. This page will hold both source tables. Add two additional pages: Category Cost for the Row to Column result, and Delay Analysis for the Column to Row result.

Add the PLUGS source table

On the Data page, add a new table element and connect it to PLUGS_ELECTRONICS_HANDS_ON_LAB_DATA from the RETAIL schema. Leave the table as-is — no grouping or filtering needed before applying Transpose:

Create a child table to transpose

Open the 3-dot menu on the table and select to create Child table:

Using the 3-dot menu but from the child table, move it to the Category Cost page.

Transpose the child table

On the Category Cost page with the child table selected, click the 3-dot menu and then select Element source > Transpose:

The Transpose configuration modal opens. The right panel labeled Original table shows a preview of the source data.

Configure Row to Column

The modal defaults to Row to Column. Leave that selected and configure the following fields:

Sigma reads the unique values in Product Type and creates one output column for each — Arts & Entertainment, Computers, Entertainment, Mobiles, Music, and Photography. We will use these in a moment.

The Value column (Cost) provides the numbers that populate those columns, aggregated by Sum within each output row.

Select output columns

In the Output columns section, choose which additional columns to include alongside the transposed category columns. Check Store Region and Store State to retain the store location dimensions.

The output preview shows the row structure — one row per Store Region and Store State — but the Product Type columns are not included yet.

Scroll the Output columns list to select the product types to include. Select them all:

Click Submit.

After formatting the values as Currency, the table now shows store locations as rows and product categories as columns, with total cost summed across all transactions in each cell.

This layout makes it straightforward to compare cost patterns across categories for any store region or state:

We could also take the extra step to group by Store Region and Store State to make the results really clear:

WHY IT MATTERS:
 The Row to Column direction eliminates the need to manually pivot or restructure data in the warehouse to produce a wide comparison layout. The reshape happens in Sigma, on live data, without any SQL or ETL work.

Footer

The Column to Row direction converts multiple columns into a single long-format output — each selected column becomes a row, with a label column identifying its origin. Use this when your source data has related metrics spread across separate columns that should be treated as a single measure.

The data:
FLIGHTS from Sigma Sample Database > EXAMPLES > FAA — a flight operations dataset including Airline and five delay columns: Air System Delay, Security Delay, Airline Delay, Late Aircraft Delay, and Weather Delay.

The problem:
 Each delay type is a separate column. To compare delay types across airlines in a grouped view or chart, all delay values need to be in a single column with a label identifying the type.

Add the FLIGHTS source table

On the Data page, add a second table element and connect it to FLIGHTS from Sigma Sample Database > EXAMPLES > FAA:

The FLIGHTS table has five delay columns — Air System Delay, Security Delay, Airline Delay, Late Aircraft Delay, and Weather Delay — each representing a different cause of delay in minutes.

Create a child table

Once again, create a child table from FLIGHTS and move it to the Delay Analysis page.

Add a Transpose table

On the Delay Analysis page, with the child table selected, click the More menu (three-dot icon) on the element, then select Element source > Transpose.

Configure Column to Row

In the modal, select Column to Row.

The configuration fields change to reflect the column-to-row direction.

For Columns to merge select the five delay columns:

Change the label for Column label for merged columns to Delay Type.

Also change the label for Column label for values to Delay Time:

The Columns to merge selection determines which columns are collapsed into rows.

Column label for merged columns names the new column that will hold the original column names as values.

Column label for values names the column that holds the corresponding numeric values.

In the Output columns section, ensure Airline is included. Leave all other columns selected.

The output preview shows the long-format result — each flight now appears once per delay type:

Click Submit.

Group within the Transpose element

Sigma creates the Transpose table on the Data page and names it Transpose of FLIGHTS.

Now we can add grouping and a calculation to summarize the data.

In the element panel, drag Airline and Delay Type to the Group By section. Then click + next to CALCULATIONS and enter:

Sum([Delay Time])

Rename the column Total Delay.

The table now shows one row per airline per delay type — 70 rows for 14 airlines × 5 delay types — with total delay in minutes.

Each airline's total delay minutes broken down by cause, in a format that supports filtering, sorting, and charting with a single value column.

WHY IT MATTERS:
 When delay types — or any set of related metrics — are stored as separate columns, comparing them requires either pivoting the query or working around the wide layout in every chart and filter. Column to Row collapses that complexity in one step, producing a long-format table that works naturally with Sigma's grouping and visualization tools.

Footer

Transpose is a powerful reshaping tool, but its behavior differs from standard Sigma child table operations in important ways. Understanding these constraints before you build will save debugging time.

Transpose operates on source data, not Sigma-computed results

Transpose goes directly to the data from your data platform — it does not operate on groupings, filters, or calculated columns applied to a Sigma workbook element. If you apply grouping or filtering to a table and then add Transpose, those transformations are not reflected in the transposed output.

This is by design and documented behavior. It is the reason this QuickStart adds Transpose directly to source tables rather than to grouped child tables.

Grouping and aggregation can be added within the Transpose element after creation — as shown in the Column to Row example — but they cannot be inherited from a parent element upstream.

Input table restriction

Transpose cannot be applied directly to a Sigma Input table. If your source data is in an Input table, add a child table first and apply Transpose to the child.

Row to Column: single value column only

The Row to Column direction supports only one value column per Transpose table. If you need to transpose multiple metrics into columns simultaneously, either use multiple Transpose tables or restructure the source data upstream.

Row to Column: 200-column limit

The Row to Column direction generates one output column per unique value in the label column. If the label column contains more than 200 distinct values, the transpose will generate an error. Design your source data to limit cardinality before transposing if this is a concern.

Row to Column: static column set

The set of output columns is determined at the time the Transpose is created and does not dynamically update when the connected data source changes. If new values appear in the label column after the Transpose is configured, the output will not include them — the Transpose must be reconfigured to reflect new values.

This limitation does not apply to Column to Row — output rows are derived from the selected source columns, which are fixed by design.

For the most recent information, see Limitations

Footer

This QuickStart demonstrated both directions of Sigma's Transpose table using two real datasets — Row to Column with PLUGS Electronics for a wide category cost comparison, and Column to Row with FLIGHTS to convert five delay columns into a long-format breakdown by airline and delay type.

The key architectural point is that Transpose operates on data from the source, not on Sigma-computed transformations. Grouping, filtering, and calculations applied to a parent workbook element are bypassed. The correct pattern is to add Transpose directly to the source table, then add grouping and calculations within the Transpose element itself — as the FLIGHTS example demonstrates.

Both directions solve the same underlying problem: data that arrives in the wrong shape for the analysis you need.

Knowing which direction to apply — and understanding that the reshape happens at the source layer — is what makes Transpose predictable to use.

Additional Resource Links

Transpose table
Blog
Community
Help Center
QuickStarts

Be sure to check out all the latest developments at Sigma's First Friday Feature page!

   

Footer