DATASET

Generates a table row by row, where each row is computed from the previous one. DATASET takes a row count followed by one or more column definitions, each with a starting value and a lambda that calculates the next value — useful for building sequences, running totals, and iterative calculations without writing each row by hand.

Syntax

DATASET(
  rowCount,
  currentName1, initialValue1, nextName1, lambda1,
  currentName2, initialValue2, nextName2, lambda2,
  ...
)

rowCount (Number): Number of rows to generate. Must be at least 0.
currentName (String): Name under which the value from the previous row is available.
initialValue: The starting value used for the first row.
nextName (String): Name under which the newly computed value becomes available to later columns in the same row.
lambda (Function): Function that computes the next value.

Returns: Table – a table with one row per generated step. Returns an empty table when rowCount is 0.

Requires at least 5 arguments: the row count plus one complete column block of four arguments. Each additional column adds another block of four.

How it works

The first row contains the initialValue of each column. For every following row, the previous row's values are exposed through the currentName names, and each column's lambda computes the new value. Columns are evaluated left to right, so a later column can read a newly computed value from an earlier column in the same row through its nextName.

Names like "x" and "x2" are strings and must be quoted. Lambda parameters are identifiers and are not quoted.

Examples

Build a simple sequence

DATASET(3, "x", 1, "x2", function(x: x + 1))

Returns the table [[1], [2], [3]].

Two dependent columns

DATASET(
  3,
  "x", 1, "x2", function(x, y: x + 1),
  "y", 10, "y2", function(x, y, x2: x2 * 10)
)

Returns the table [[1, 10], [2, 20], [3, 30]]. The second column reads the first column's newly computed value x2 from the same row.

Running total

Accumulate a running total across rows by adding each row's previous value:

DATASET(4, "total", 0, "total2", function(total: total + 100))

Returns [[0], [100], [200], [300]].

Common mistakes

Unquoted names — currentName and nextName must be strings ("x"). Unquoted x is treated as a variable reference.
Wrong argument count — each column needs exactly four arguments (currentName, initialValue, nextName, lambda).
Fractional row count — rowCount is rounded down with Math.floor. A value like 3.9 generates 3 rows, and 0.9 generates 0 rows.
Lambda errors — if a lambda fails while building a row, DATASET passes the error through and does not add the incomplete row.

DATASET runs in the browser, but the server-side evaluation returns an error because it cannot execute lambdas. Formulas that rely on DATASET may therefore fail in server-side contexts.

Error messages

Error	Cause
`DATASET expects at least 5 arguments`	Missing the row count or a full column block
`DATASET expects a number as the first argument`	Row count is not a number
`DATASET expects a row count of at least 0`	Row count is negative
`DATASET expects columns in sets of 4 (currentName, initialValue, nextName, lambda)`	Arguments after the row count are not grouped in blocks of four
`DATASET expects current name to be a string`	`currentName` is not a string
`DATASET expects next name to be a string`	`nextName` is not a string
`DATASET expects lambda to be a function`	A lambda argument is not a function

DATASET

Syntax

How it works

Examples

Build a simple sequence

Two dependent columns

Running total

Common mistakes

Error messages

See Also