Module 3

Power BI
It maintains 2 interfaces – service website and desktop.

Service interface:
 There is a service tab where you can find your recent elements,
create new elements, and find elements shared with the current
user.
 There are also training workspaces, my workspace and get data.
1. Get data is what you use to begin the process of creating a report
by connecting to outside data sources or accessing data in the
Microsoft Dataverse.
2. Workspaces are places where the individual users elements are
stored.
 PBI has 2 views – reading (where we can see the report and use
filters) and edit view (to create changes).

Reading view Edit View

Desktop App:
 Windows application.
 The Power BI desktop is strictly a development tool. It does not
provide data to users directly. In order to do that, you have to
publish the report or the dashboard to the service, which makes it
accessible from the cloud.
 There is a ribbon at the top of the screen, not unlike the ones that
appear in office applications.
1. The report tab displays the pages of the current file and the
visualizations and filters panes which users can employ to create
and modify those pages.
 2. The second tab is data, which displays the contents of the
documents data sources and the controls for adding measures
and transformations.
3. And finally is the model tab which is used to display the
relationships among tables when there are multiple data sources
in a report.
 The advantage of Power BI Desktop is that it contains much more
powerful data cleaning, transportation, and manipulation controls.
So you can work with your data more precisely in Power BI desktop.

Dashboards
 To create a dashboard we need to create a report first.
 A dashboard is similar in appearance to a report, particularly since a
dashboard is created from a report.
 But there are no pages, a dashboard is only one page, and there are
no filters. You can't apply filters to a dashboard although the
developer can apply them when creating the dashboard.
 Basically, a dashboard is a means to display live data to users
without the need for them to modify it heavily. There's a little
question and answer space where a user could ask a question of the
developer and the dashboard is therefore shareable amongst users
without them needing to do anything in order to view the data.
 A dashboard is only a single page, so developers must consider the
size of the screen that consumers will be using to view it.
 Dashboards are scrollable vertically and horizontally, so you can
conceivably create a page that's longer or wider than the actual
screen.
 After a developer has pinned all of the tiles to a dashboard that they
want to appear there, it's possible to move them around, resize
them, and add titles and subtitles to create a usable configuration.
 The choice of the display is up to the consumer, but developers
might want to consider sizing the dashboard to accommodate full
screen mode in which the dashboard appears without any
surrounding frames and navigation panes so their screen is only
 displaying the dashboard itself. So design your screens for that size
and you can maximize the amount of space that's available to your
consumer.

Workspace
 Power BI contains workspaces in the service interface. A workspace
is simply an area where the current user's content elements are
stored and where they can access them from anywhere because
they're in the cloud.
 There are three basic projects here with several different elements
in each one. Clicking one of these items in the list will open it in the
Power BI service interface.
 If desired, a user could create a workspace for each product that
they're working on or divide their projects in any other means that
they want to do.
 If you are working with different teams in the organization, you
might want to have one workspace accessible to team A while
another workspace is accessible to team B.

Visualization controls
 These different types of graphs are called visualizations in Power BI.
 The visualizations pane is where you select the type of graph you
want to display in a particular tile.
 The dashboard is zoomed in on one particular tile and the
visualizations page when that tile is selected. You can modify that
graph if you select a different visualization.
 Bar chart: The most basic and the most common on that
dashboard are bar charts and column charts.
 Stacked Column: This is a stacked column chart which you'll see
has two values in each of the columns.
This enables you to compare the values within each column as well
as the values between columns.
A 100% stacked column chart expresses data in the same way,
except that it uses percentages instead of actual numerical values.
 Clustered Column chart: A clustered column chart simply groups
the columns together so that like values can be seen next to each
other.
 Funnel Chart: A funnel chart is essentially a bar chart that displays
values in descending order, creating something that looks like a
funnel.
 Waterfall: A waterfall chart divides a single statistic into a group of
sequential increments.
  Line Chart: A line chart is very common. Usually seen in terms of
time lapses where the x axis along the bottom of the chart specifies
months or years and the y axis specifies some numerical value.
 Combination: A combo chart, as the name implies, is a
combination of two or more types of charts.
 Area: An area chart is simply a line chart that has the area below
the lines shaded all the way down to the x axis. By making the
shading translucent, you could also see the interface between the
two data sets.
 Pie: A pie chart is a circle divided into wedges, usually employed to
indicate percentages of a whole.
 Scatter: A scatter chart does not have a time scale on the x axis, it
has some other numerical or other data value.
 KPI: The key performance indicator chart specifies the progress of a
single data point toward a specific goal which is specified below the
current point in the middle of the chart.
 Card: A card is just a tile with numerical data in it.
 Gauge: A gauge chart uses the paradigm of a fuel gauge to indicate
the progress towards a specific goal.
 Key Influencer tile: A key influencer tile is a little different in that
it uses some intelligence from Power BI to analyze the data in
different fields in the data set. So when you drag and drop fields into
the left pane of this tile, the tile creates the chart on the right and
can manipulate it based on how the user interacts with the various
fields listed there.
 Regular table
 It is also possible to create customized visualizations in Power BI
which could have various applications.
 The AppSource Store contains a variety of visualizations created by
Microsoft or by other users that have been submitted to Microsoft.
So there are quite a lot of different examples of visualizations
available there for your exploration.

Types of data sources

 In the Power BI service there four tiles that enable you to access
different kinds of data.
 The My organization title is to access data that's been generated in
house and published by other people in your organization.
 The services tile enables you to choose apps from online services.
The files tile enables you to open files stored locally or on the
network such as Excel spreadsheets or Power BI Desktop files.
Incidentally, this is how you would get your work from Power BI
Desktop into the Power BI service.
  And then finally you have the databases tile which enables you to
access Azure SQL server or local server databases.
 When you click one of these and select a data source, it might be
supported directly in the Power BI service, but some of them as in
the Azure SQL database tile selected here, is telling the Power BI
service user that in order to connect to this particular database, you
need to use Power BI Desktop and it provides a link to download the
desktop application.
 Power BI Desktop has access to all of the available connectors and
data sources while Power BI service is relatively limited in that
respect.

Power Query to clean data

 The purpose of the data modeling is to select and arrange the data
that you've accessed from the data source so that it suits the
visualization the developer intends to create.
 The Power Query editor is available only in Power BI Desktop. It's not
part of the service interface.
 Right click to find options to transform the type of data for example
from text to numbers. We can also delete any non required columns
by right clicking and selecting delete.
 There's a pane called applied steps, which lists all the things you've
done to the data in Power Query editor and you can easily undo
anything you've done by simply deleting that step with the little X
there.
 To combine 2 tables, select merge query under combine. When you
merge queries, you're essentially adding columns from one table to
the other table. I could also choose append queries, which doesn't
apply in this case, but appending queries enables you to add rows
from one table to the other table.
 The merge function will corelate data even if the data is in different
orders in different tables if there is a column that will serve as a
basis for that corelation. (Eg. If there is a column with state codes
and state names in table 1 and only state names in table 2, then
using the state column, PBI will merge the 2 tables correctly.)
 After clicking OK we see we have emerged query. There is a new
column here and it has table as the value. And that's indicating to
us it's a placeholder for the actual table that we've imported. By
clicking the expand icon at the right of the column heading, you get
this dialog box in which you could select which columns from the
merged table you want to appear in your target table.
 Aggregates are another feature of Power BI, which appears at times
when you import specific types of data automatically. Power BI
automatically summed up the units sold from all of those countries
 and put them all together into one total for each product. But, you
don't have to keep it that way. You can right click the visualization
and you see here's what's happening with that aggregation and sum
is selected. I could change it to an average, to a minimum or
maximum or any of these other options.

 You can create a shared dataset and give others access to it. You
can also decide how much access that person receives.
 Endorsements - Promoted label in a blue box indicates that this is a
data set that's been specified by internal users as recommended.
The certified tag indicates that the data set was submitted to
Microsoft for certification.
 Another tool that can aid users in creating their dashboards and
reports is to open a template app. One of the advantages of this is
that you could take an existing app and connect your own data to it
so that the app's already in place, you're just modifying the data
that it displays. When you select an app like this, it creates a new
workspace in your account.

Sharing Reports and Dashboards

 Power BI was designed for developers to be able to create content
and share it with consumers, and they can do that in several
different ways.
 They can publish a Power BI dashboard or report to the Power BI
service, and then by sending an email link to consumers, they can
access the content from this service.
 Developers can also build apps that contain multiple reports and
dashboards into a single unit and share it with consumers by
enabling them to install the app in their workspaces or by sending
them an installation link.
 Another method is to export the Power BI content. They can save
the report or dashboard as PDF file or a PowerPoint slide deck or it
can simply be printed or exported to Excel.
 Now consumers can also interact with Power BI content. For
example, they can subscribe to a dashboard which would cause
Power BI to send them emails when content changes so they could
be regularly notified when new content appears. Another method is
alerts.
 Consumers can create alerts using this interface which causes
Power BI to send them an email when the value for a specific data
point reaches a certain threshold.
 Also users can interact with the content by adding comments to the
actual dashboard or report. And there can actually be a dialogue
 going there between various users in this comment field as you see
it here.

BI Reports
Process –
Connecting to a data source -> click create a report -> create
visualizations -> and then sharing the report.

BI Dashboards
Dashboards are essentially gateways to the reports where the tiles
originated. Clicking one of the tiles opens the report, assuming it has been
published.
Dashboards have advantages and disadvantages.
They're limited to a single page, which means that the amount of
information is relatively limited. But that can be considered an advantage
as well, because great many consumers do not need the full range of
detail that they might find in a multi page report.
The dashboard contains only the report elements that the designer
determines the user needs.
Another advantage of dashboards is that you can take tiles from multiple
reports and data sets and insert them into one dashboard. So if you have
content that's distributed among a number of reports, you can create a
consolidated view of the most important information by creating a
dashboard.
Dashboards are limited to the Power BI service. You cannot create a
dashboard in Power BI desktop.
So the easiest way to create a dashboard is to start with a report and
choose one of the visualizations that you want to appear on the
dashboard.
Every tile in a report has a series of icons in the upper right corner which
appear and disappear according to where your mouse is. But that toolbar
is where you begin this process.
When you click the push pin icon in the toolbar, the pin to dashboard
dialog box appears as shown here. You can select whether to create a new
dashboard, with this tile, or add it to an existing dashboard.
It's also possible to add your own tiles that contain other types of content.
In the dashboard Edit menu if you select the add a tile button, you get an
add tile dialog box in which you can choose to add a tile containing web
content, an image file, video, a text or custom streaming data.
It's also possible in the Power BI service to view a dashboard in a mobile
format, assuming that it's intended for use on cell phones. This is an
important factor that the developer has to consider when creating a
dashboard. These dashboards might need the chart types to be changes
to make them readable (Like I had to do for a dynamic website, except its
not dynamic for dashboards, we need to choose 1 mode only).

Publishing and Sharing Reports

 When you create content with Power BI, it's not accessible to
anyone else unless you publish it to the Power BI service.
 To share it, you specify the name of a user or an email address of a
user. It does not have to be a user in your organization, you could
share with outside people.
 There are 3 options for the kind of access you want to provide –
1. The first checkbox, allow recipients to share your dashboard
report, means that the recipients of access can share the
dashboard with other people if you want them to. Obviously, if
you had sensitive data that you needed to protect, you would
clear this checkbox.
2. The second one is allow recipients to build new content using the
underlying data sets. If you select this checkbox, then recipients
have access to the data sets you use to create the report or
dashboard, and they can use that data themselves to create their
own reports or dashboards, even using the developer's
credentials in the process. Obviously for sensitive data, this is
something you have to be very careful about.
3. And then finally, you could choose to send an email notification
to recipients informing them that they have been granted access
to a particular report or dashboard.

 On Power BI Desktop, you have to publish any content you create to

the Power BI service before you can share it.
 So in Power BI desktop, you would be working with the report
because you can't work with dashboards. You would open a report
and click the publish button on the home tab and or select from the
menus file, publish to Power BI. And a dialog box appears which lists
the workspaces that you can use as a target for the material you're
publishing.