Luyện thi Tableau Desktop Specialist

Tableau Desktop is a self-service visual analytics tool for connecting to data, exploring it, and building interactive dashboards. It is not designed for raw database editing, app development, or model training.

A .twbx packaged workbook stores worksheets, dashboards, stories, and supporting data such as extracts. A .twb workbook stores only the workbook definition, while .hyper and .tde are extract formats.

In Tableau, blue field icons represent dimensions, which are usually discrete fields that create headers and group data. Green icons represent measures, which are usually continuous and aggregated.

Green icons identify measures, which are typically numeric fields that can be aggregated with SUM, AVG, COUNT, and other functions. Blue icons identify dimensions used for grouping and headers.

The Rows shelf controls the vertical axis, and the Columns shelf controls the horizontal axis. Filters limit data, and the Marks card controls the appearance of marks.

The Data pane shows dimensions and measures from the connected data source. The Analytics pane adds reference lines and forecasts, while Show Me suggests chart types.

An extract is a local snapshot of data stored in the .hyper format. Extracts improve performance and allow offline work, but they must be refreshed to reflect changes in the source.

A live connection queries the database in real time, so it is best when data must stay current. Extracts are better for offline work and performance, but they contain a snapshot.

A packaged .twbx file bundles supporting resources such as extracts, custom images, and fonts. A regular .twb file stores only the workbook definition and relies on external connections.

A pie chart divides a whole into proportional slices, which makes it easy to compare category shares. Scatter plots show relationships, box plots show distribution, and Gantt charts show schedules.

The Microsoft Excel connector reads .xlsx and .xls workbooks directly. JSON, MySQL, and OData are connectors for other data source types.

An extract creates a local snapshot in .hyper format, allowing work without a live database connection. Live connections require access to the source and always query it.

The Data Interpreter scans spreadsheet or text files for header rows and extra metadata, then prepares a clean table for analysis. It does not write to the source or change file formats.

Tableau has native connectors for cloud services such as Google BigQuery, Amazon Redshift, Snowflake, and Azure Synapse. Local text files are not cloud sources, and PDFs and screenshots are not direct data connectors.

Refreshing an extract pulls the latest data from the source into the local .hyper snapshot. The extract remains a snapshot and must be refreshed again when the source changes.

Tableau Prep Builder provides a visual interface for cleaning, combining, and reshaping data. Tableau Public shares visualizations, Tableau Bridge supports on-premises connections, and Tableau Mobile is for viewing content.

A join combines rows from two tables by matching values in common fields. A union stacks tables with the same structure, a pivot reshapes columns into rows, and a split divides one field into multiple fields.

A union appends rows from one table below another when the tables have matching columns. A join combines columns, a filter removes rows, and a group combines values.

A filter step keeps only rows that satisfy the condition, which is useful for removing incomplete records or limiting data. Split, aggregate, and sort do not remove rows by condition.

A pivot step transforms wide data by moving selected columns into rows, creating a longer and more analysis-friendly structure. It does not join, type-cast, or rotate the interface.

Changing a field data type from string to date lets Tableau recognize it as temporal data. Split and union restructure data, while joins combine tables.

A split operation divides a field into multiple fields based on a delimiter, such as splitting a full name into first and last names. Joins combine tables, null cleaning replaces missing values, and dashboards are separate objects.

Assigning a geographic role such as Country or State tells Tableau how to interpret the field for maps. Numeric, date, and boolean roles do not provide map coordinates.

Replacing nulls fills missing values with a chosen value so downstream calculations behave predictably. Split, pivot, and union change the structure rather than fixing missing data.

A calculated field uses a formula to derive a new value from existing fields. Reference lines and tooltips appear in visualizations, and dashboard actions create interactivity.

A wildcard union combines files whose names match a pattern, such as sales_2026_01.csv, sales_2026_02.csv, and sales_2026_03.csv. It is useful for regularly named files.

A line chart connects points over time and makes direction and rate of change easy to read. Pie and tree maps show proportions, while Gantt charts show project schedules.

A scatter plot places one measure on each axis and reveals correlation or outliers. Pie charts show shares, bar charts compare categories, and histograms show distribution of one measure.

The Size option on the Marks card controls the visual size of marks, which can also encode a measure. Color encodes categories or values, Label adds text, and Tooltip controls hover details.

The Analytics pane offers reference lines, band lines, trend lines, forecasts, and other statistical objects. The Data pane lists fields, and the Marks card controls appearance.

A reference line marks a constant value, an average, a median, or another statistic across the view. It helps users compare marks against a meaningful threshold.

A forecast uses historical time-series data and statistical models to project future values. Reference lines mark thresholds, highlight actions emphasize marks, and tooltips show details.

A dual-axis chart places two measures on separate axes in the same pane, often using different mark types such as bars and lines. It is a worksheet feature, not a collaboration feature.

The mark type selector on the Marks card switches between bar, line, circle, square, and other mark types. The Data pane lists fields, and refreshing updates data.

A heat map encodes density or magnitude with color intensity across a matrix. Pie charts show shares, box plots show distribution, and Pareto charts combine bars and lines.

A story arranges worksheets and dashboards in order so readers can follow a narrative. It is a workbook object, not a connection, extract, or calculation.

TODAY() returns the current date, while NOW() returns the current date and time. DATE() converts a value to a date, and Tableau does not use CURRENT_DATE().

SUM() adds the values of a measure in the current view. COUNT() counts records, AVG() calculates the average, and MIN() returns the smallest value.

Table calculations compute values using data in the current view, so results change when dimensions, filters, or sorting change. Extracts and data source filters operate on the data itself.

Percent of total divides each value by the total for the chosen scope, such as across the table or pane. Running total accumulates values, rank orders them, and difference compares neighbors.

Rank assigns values a position such as 1, 2, or 3 based on their size in the view. It does not remove data, change marks, or add reference lines.

A FIXED expression computes at the specified dimensions and ignores view dimensions. INCLUDE adds dimensions to the view level, and EXCLUDE removes a dimension from the calculation.

Tableau applies extract filters and data source filters first, then context filters, then dimension and measure filters. Context filters can improve performance and change how other filters behave.

A Top N filter keeps only the highest or lowest N values based on a measure. Context and extract filters remove data before the view, and highlight actions emphasize marks.

In Tableau, the plus sign concatenates strings in calculated fields, for example [First Name] + " " + [Last Name]. The asterisk multiplies numbers and is not a string operator.

IF and IIF evaluate a condition and return different values for true and false outcomes. ROUND rounds numbers, LEN returns string length, and TRIM removes spaces.

INCLUDE computes a value at the specified dimensions and then includes that result at the view level. It does not remove fields, create extracts, or change connections.

Measure filters apply to continuous measures and use conditions such as greater than, less than, or between. Dimension filters apply to categorical fields, and wildcard filters use patterns.

A dashboard combines worksheets, images, text, filters, and other objects on one interactive canvas. A worksheet is a single view, not a collection of views.

Tiled objects snap to a grid and move together when the dashboard resizes. Floating objects can overlap and be positioned freely on the canvas.

A filter action uses a selection in one sheet to filter marks in other sheets. Highlight actions emphasize matching marks, URL actions open links, and parameter actions update parameter values.

A highlight action keeps all marks visible but emphasizes those that match the selection. A filter action removes nonmatching marks, while extracts and joins are data operations.

A parameter is a user-facing control that stores a value, such as a threshold or top N count, and can drive calculations and filters. Reference and trend lines are analytics objects.

Tableau Public lets users publish and share visualizations publicly for free. It is not for private enterprise publishing, password storage, or dedicated ETL work.

Tableau Server and Tableau Cloud host workbooks for organizational sharing with permissions and governance. Tableau Public is public, while CSV and PDF are export formats.

Extracts store compressed data locally and often load faster than live queries on large sources. Tooltips, mark size, and chart type do not address data-source performance.

Device layouts let you design different dashboard arrangements for phone, tablet, and desktop. Tooltips, marks, and Show Me do not create responsive layout variants.

The .twb and .twbx files are Tableau workbook formats; .twbx is the packaged version that includes supporting data. The .hyper file is an extract, and .pdf is an export format, not a workbook.

An extract is a snapshot of data at a point in time. It only contains the latest values after it is refreshed, so the statement is false.

The Rows shelf controls vertical position, the Marks card controls color, size, and detail, Show Me suggests chart types, and the Data pane lists available fields.