Profiles

3 Minutes to read

Print
Dark
Light
PDF

Article summary

Did you find this summary helpful?

Thank you for your feedback

A key feature of DataJet is the ability to rapidly profile data on very large datasets without needing to create SQL queries. This gives Instant insight into the data without the need for complex engineering.

Profiles are used for:

summarizing
visualizing
counting, frequency, grouping, aggregating, cross-tabulating

All profiles count and summarize individual records in order to understand the distribution of the data:

SQL

The SQL equivalent to DataJet profiles would involve some form of GROUPBY or PIVOT clause. There is no need to create a SQL query in order to generate profiles in DataJet. All profile types can be generated by dragging data objects from the Project Explorer and dropping them onto a report or dashboard, or by right-clicking the object in the Database tree and choosing a profile option.

As well as simple instant profiles, sophisticated calculations can be configured to group and display multiple measures. It is also possible to visualize relative under or over-representation of unique values in different datasets, and to create calculated columns from profile results. Profile result sets can be output to file, saved as new tables, used as input for models and added to dashboards and standardized reports. Underlying data can be selected from grids and graphs and used to create datasets for further analysis.

Profiles can be rapidly duplicated with new dimensions or filters by dropping field templates onto an individual profile report. Variables can be used to automate the production of all profile types in script, and many different chart types and layout options are available.

All categorical data is suitable for profiling - continuous data can be engineered so that it is suitable for profiling.

Types of Profile

The following profile reports are available:

Profile Type	Example	Used For
Profile		Use for instant counts and summaries. Accessed by Database Tree right-click or from Analytics \| Profile Simple Profile - distribution grouped by a single dimension(e.g. count of Gender) 2-dimensional Profile - distribution grouped by 2 dimensions (e.g., count of Gender by Marital Status) Function Profile - single with a function applied (eg., sum of income grouped by Gender) Function Profile - multiple dimension with a function applied (eg., sum of income grouped by Gender and Marital Status) Crosstab - 2 dimensions, pivoted by one dimension
Index Profile		Used to determine under and over-representation of values in datasets. Accessed by Database Tree right-click or from Analytics \| Index Profile 1-Dimension - under or over representation of values from a single dimension 2-Dimension - under or over representation of values from 2 dimensions
Multi-Function Profile		Used for summaries and visualisations where multiple measures and/or further manipulation of results is required post-processing Simple Profile - one or more dimensions Multi-Function Profile - one or more measures Multi-Segment - one or more segments
Combined Profile		Used to join one or more simple profiles sharing common categorical values into a single result set. Allows data from different tables to be seen in the same visualisation Simple Profile - one or more profiles
Field Profile		Create individual profiles for one or more fields, showing the count of records for each discrete value in each field.
Query Matrix		Analyze non-exclusive or disconnected segments in the same report. Used to group segments not directly linked in the Data Model.