Follow the example work flow to familiarize yourself with using "Magellan" to improve
your business.
Procedure
- Log in to "Magellan". Your credentials are managed via SAP Identity because "Magellan" is a
SAP HANA Cloud Platform (HCP) project application. When you are logged in, you
are brought to an exploration home screen that has samples and previously built
explorations. Explorations are the primary structure that users interact with in
"Magellan". Each exploration consists of a dataset and all of the analytics
associated with the dataset.
- Create a new exploration (or delete or move existing explorations). To create a new
exploration, you upload a wide dataset in a CSV file into "Magellan". When the
file is uploaded, "Magellan" displays all of the fields that are included in the
dataset, as well as information about the dataset size and scope.
Note To ensure that you upload suitable data, read the section, How do you achieve accurate
results?
- Define the business question using the following process:
- Choose the part of the data that you want to explore. The part must be present in your
dataset and can be a field, multiple fields or a range of values. This
is known as the target. For example, you might want to increase revenue.
In this case, you must know the factors that influence revenue and to
explore the best way to increase it. Therefore, you would click the
Revenue field. This action asks the application to build a data model
based on the revenue field.
- Examine the data that "Magellan" returns. The application runs Advanced Analytic
algorithms over the data against the target and displays Contributing
Variables, which show the data that has the most impactful relationship
with the target. The key influencers display prominently also. In the
case of a revenue exploration, a key influencer could be the amount of
services sold. What’s more, the accuracy of the model and its robustness
in the event of receiving additional data is given an Insight Quality
Value.
This is "Magellan" main screen which displays key influencers for the revenue field of a
Travel Booking Dataset exploration:
- Interact with the output to see if anything interesting has emerged. Select the circled
Contributing Variables to see the nature of their relationship with the
target across a number of Variable Categories. Variable Categories are
columns that are grouped together because they behave the same; for
example, the dataset columns Hotels and Excursions can be grouped
together as they attract tourists.
- Explore and discover insights. Delve deeper into the data by running more analyses
against interesting parts of the information. For example, you can run
an analysis on two related influencers to understand their impact. The
results are visualized in various charts and maps, including a data heat
map to view a range of numbers.
- Create a What-If Analysis. This enables you to adjust influencer variables to observe the
impact on your business. Traditional analytical tools that comprise basic
business intelligence (BI) examine historical data, whereas tools for advanced
and exploratory analytics focus on forecasting future events and behaviors,
which enables businesses to conduct what-if analyses to predict the effects of
potential changes in business strategies.
- When you end your session, "Magellan" stores the information with which you were working
until it is explicitly deleted.