Module 3
Calculation to Enhance Data
Aggregation in Charts
Tableau uses aggregation to denote how separate records of data are summarised when they
are introduced into a visualisation. Tableau will automatically summarize the measures based
on the detail level on the dimensions used in the view. As an illustration, when a dimension
like Department is added to the view and a measure like Patients Handled is added, Tableau
will sum up the number of patients handled by each department. The reason behind
aggregation is that summarised data is often used to make most managerial decisions as
opposed to raw transactional records. It enables the analyst to learn trends, draw comparisons
among groups, and eliminate noises.
In real-life situations, aggregation is applied in performance dashboard, workload analysis,
financial summaries and operational reporting. Unless aggregated correctly, charts may be
confusing. Knowledge of aggregation also assists the analysts to realise that values transform
as dimensions are added or removed to a view.
Steps in Tableau:
In order to apply aggregation, drag one measure in the view, and one dimension. Tableau is
automated in aggregating the measure (which is typically SUM). To check the aggregation,
go to Analysis - Aggregate Measures and make sure it is checked. The aggregation may be
altered, by clicking on the drop-down arrow on the measure pill, and choosing the required
type of aggregation.
Aggregation Types
Types of aggregations describe the form of summarisation of values. Tableau allows some
aggregation functions, which include: SUM, AVG, MIN, MAX, COUNT, COUNTd and
MEDIAN. The types of aggregations have varying purposes of analysis. SUM is suitable
when the total such as revenue or workload is involved whereas AVG is applicable when the
rating, response time, or satisfaction level is involved. COUNT is an operator used to count
records whereas Countd is an operator used to count unique values like the number of
customers or employees.
Making a wrong decision on the type of aggregation may end up giving a wrong conclusion,
including adding satisfaction scores rather than averaging them. Hence, the knowledge of
types of aggregation is essential in the accuracy of analysis.
Steps in Tableau:
On the measure in the view - right-click (or click on the drop-down arrow) - Measure - pick
the aggregation you want (SUM, AVG, COUNT etc.). Tableau re-calculates the chart
promptly depending on the choice of aggregation.
Table Calculations and Calculated Values.
Computed values (calculated fields) enable analysts to write their own new fields using
formulas on existing ones. These computations are computed at the data level and are not
dependent on the manner in which the view is constructed. Table calculations, in their turn,
work with aggregated outcomes and are based on the structure of the view. Profit margin, on
the other hand, is a calculated value, whereas running total of sales is a table calculation.
Derived metrics such as ratios, percentages, flags or categories are done in calculated fields.
Table calculations are employed in cases where analysis needs to be compared either by
rows, columns or time.
Steps for Calculated Fields:
Go to Data Pane - Right-Click - Create Calculated Field, type formula with Tableau
functions, assign it a name and click OK. The new field is presented in the data pane.
Table Calculations Steps:
Add a measure to the view - choose the measure pill drop-down - choose Quick Table
Calculation or Add Table Calculation.
Formula with Table Calculation.
Table calculation formulas can be used to perform more sophisticated analysis like
cumulative totals, moving averages, ranking and growth rates. These computations act on
aggregated information that is already in the view. Indicatively, such things as running total
can represent cumulative performance over time, whereas the RANK can be used to
determine high-performing categories.
These calculations are particularly applicable to trend analysis, performance monitoring and
contribution analysis where relative movement has a stronger role than absolute values.
Steps in Tableau:
Add a measure to a view - click the drop-down on the measure - add Table Calculation - pick
the calculation that is needed like Running Total, Percent of Total, Difference or Rank.
Tableau implements the calculation on the basis of the existing view structure.
Table Calculation Customization.
Table calculation customisation entails the management of calculation in the view. Table
calculations require performing calculations to be run across a row or down a column or
specific dimensions; therefore, analysts need to state how they want the calculation to be
done. Improper customisation may lead to false intuitions.
Multi-dimensional views, like comparing the trend of a department across time or a region
across category, are important in terms of customisation.
Steps in Tableau:
Click the pill of table calculation - choose Edit Table Calculation - choose the direction of
calculation Table Across or Table Down, Pane Across or Specific Dimensions. Choose or
unchoose dimensions to regulate the scope of calculation.
The parameters are calculated and presented as follows:
Parameters are pre-defined inputs that enable the interactivity of calculations. They are not
bound to the dataset and can reflect values like targets, thresholds or scenario inputs.
Parameters allow the simulation of what-if analysis, sensitivity analysis and decision
simulation.
As an example, the management can modify a target value and immediately find out which
departments are achieving the target and which are not.
Steps in Tableau:
Data Pane - right-click - Create parameter - define data type, range and default value - click
okay. Then make a calculated field with the reference to the parameter. To make the
parameter interactive, right-click and choose Show Parameter.
Advanced Sets
Sets are subsets of data like those specified by conditions or by the user. Advanced sets
enable dynamic segmentation to include top performers, bottom performers or entities that
satisfy a set criterion. The sets have extensive applications in benchmarking, comparative
analysis and exception reporting.
In contrast to filters, sets do not lose context because they do allow one to compare between
the selected and non-selected data.
Steps in Tableau:
Multi-step process Right-Click a dimension - Create - Set - define conditions (Top, By Field,
or Formula) - click OK. Drag the set to the view to make it a dimension or filter.
Parameters with Filters
The integration of parameters and filters enable an analyst to perform filtering logic on
demand. Parameters rather than fixed filters provide the user with the opportunity to draw
thresholds or limits. It can be applied in such situations as the presentation of the top N
objects or the arrangement of data above a user-specified value.
This method increases the flexibility without altering the source of data.
Steps in Tableau:
Design a parameter - make a calculated field based on the parameter - locate the calculated
field on the Filters shelf - display the parameter control.
Advanced Sets and Filtered Parameters.
Advanced sets and parameters combined with filters make it possible to provide advanced
interactive analytics. Parameters specify conditions, sets are dynamic classifications of data,
and filters are visibility. This mix is applied in complicated business situations like
determining the units which have the greatest risk, best performers, or even the selected
entities to the rest.
The technique facilitates dynamic decision making and narrative development.
Steps in Tableau:
Form a parameter - develop a calculated field that points to the parameter - consume the
calculated field within a set - utilize the set as a filter or dimension in the perspective.
The general Significance of Calculations in Tableau.
The use of computations brings Tableau to the level of decision support system rather than a
visualisation tool. They permit analysts to provide answers to more profound questions of
trends, comparisons, contributions and scenarios. Calculation skills allow users to make
flexible, interactive, and analytically rich visualisations to aid managerial decisions in such
areas as operations, finance, human resources, and policy analysis.