Salesforce Tableau Data Analyst cheat sheet

A union appends rows from multiple tables; Tableau aligns fields by position or name, and each input must have the same number of fields.

A join combines columns using one or more key fields; many-to-many keys duplicate rows, while relationships preserve each table’s granularity and avoid unmatched-row duplication.

The Data Interpreter removes extra headers, footers, and merged cells from Excel, text, and Google Sheets sources before Tableau reads the table.

Pivoting converts crosstab columns into rows so that values become a single field; Tableau Desktop supports pivot on the Data Source page.

Extract filters are applied before the extract is created, reducing the number of rows saved in the .hyper file.

ZN(expression) returns zero for null input and otherwise returns the expression; ISNULL(expression) returns a Boolean indicating null status.

A field’s data type determines valid operations: strings support concatenation, dates support date arithmetic, and numeric fields support aggregation.

Tableau Prep flows commonly use Input, Clean, Aggregate, Pivot, Join or Union, and Output steps; each step transforms data before publishing or extracting.

Dimensions are categorical fields that partition data and appear blue by default; measures are numeric fields that aggregate and appear green by default.

FIXED level-of-detail calculations compute at specified dimensions independent of the view’s dimensions, most dimension filters, and aggregation.

INCLUDE adds dimensions to the view’s level of detail, while EXCLUDE removes dimensions from the view’s level of detail before aggregation.

Filter order is extract, data source, context, dimension, measure, and table calculation; context filters execute before FIXED LOD calculations.

A table calculation uses the current view’s data and requires Compute Using settings to define partitioning and addressing.

A parameter is a user-controlled value that can replace constants in calculations, filters, reference lines, and titles.

Bins group a continuous measure into equal-width ranges, while groups combine dimension members into named categories.

A set defines a subset of dimension members and can be used in calculations, filters, and color encodings.

The Marks card controls mark type and encodings: Color, Size, Label, Detail, Tooltip, and Path are available for relevant mark types.

A dual axis overlays two continuous axes in one view and can use different mark types or scales; a combined axis shares one scale.

Categorical color palettes distinguish discrete members, while continuous palettes encode numeric magnitude along a gradient.

Dashboard actions include Filter, Highlight, URL, Go to Sheet, and Parameter actions; each passes a selection or value to a target.

Containers group and arrange dashboard objects horizontally or vertically and control padding, distribution, and nested layout.

Device layouts store dashboard positioning for specific screen sizes, such as phone or tablet, without changing the default layout.

A viz in tooltip embeds a worksheet in a tooltip so viewers see additional detail when hovering over a mark.

A story point captures a worksheet or dashboard state with a caption, forming a sequence that guides viewers through a narrative.

A site role sets a user’s maximum capabilities on a site; common Tableau site roles include Creator, Explorer, and Viewer.

Project permissions apply to content in a project and can be set to Locked, Custom, or Managed by the project owner or administrator.

Permission capabilities include View, Filter, Download, Connect, and Publish; capabilities can be granted to users or groups.

A user filter restricts a published data source so each viewer sees only rows matching their username or group membership.

A certified data source is marked by an administrator or project lead as trusted, signaling that it meets governance standards.

A subscription emails a snapshot of a workbook or view on a schedule and is processed by Tableau Server or Cloud backgrounder processes.

An extract refresh updates a published extract on a schedule; a failed refresh can leave stale data and should be monitored by administrators.

When publishing a workbook with a live connection, credentials can be embedded or prompted so viewers can access the underlying data source.