power bi

A common way to setup simple powerbi reports


from 'Get Data' or 'New Source', select the data source information. It supports excel, csv, oledb, sql server, etc.

       it also selects the tables to be imported


       once imported, a table appears as a  query / dataset in the 'transform data' panel, from where one can modify columns, filter rows, merge tables/queryies, groupy by, pivot, etc. 

      each transformation step appears in the Applied Steps on the right hand side as history. click on each step to modify if needed.

      each step is translated to DAX script. Clicking the Advanced Editor will show the scripts

     view -> column data distribution and column profile

     Managed parameters

    In the Transform Data panel,  the Manage Parameters is only for filtering data in PowerBI desktop when developing the report. It wont be available after publishing to the workspace.

3. Table View

after Save and Close the Transformation Panel, it switchs to the report panel, which has 3 views: report view, table view and model view.

The Table view is imply a tabular display of the columns and rows.

From menu, one can also create new measure, column or table.

4. Model View

   The model view is for connecting the data tables available from the Transformation.

   Every transformed query is available as a Data table in the Model view.

   The panel shows all the tables. You can add relationship (e.g. many to 1 or 1 to 1) to between tables. This is kind of Entity Relation model.

    with the relations defined, the report view can later drag columns from multiple tables. PowerBI will do the join automatically behind based on the relationship. Another good thing is when selected a certain row in the report view, (kind of drilling into details), the other visuals based on other tables will be drilled / detailed based on the cascaded selection.

5. Report view

This is the main design panel for working on the report visulisations.

ON the left hand side,  it is the canva for arranging visuals. You can select and add visualizations from the list on the right. 

Once a visual is added, you can map the data table columns as x, y or legend, etc. depending on the visual type.

In the middle, there are filter options, this is for selecting subset of data to display.

Note that, you can add new visual from the ...

Python visual is also available, which expects a dataframe input. But only a limited number of python libraries are supported (come with the powerbi release version)

Slicer

One visual is Slicer, which is similar to filter, but you can add it as a slide bar or drop down list onto the report. 

By mapping a data table column, it can filter the report as well.

Further analyses

Right next to the visual tab, there is visual format tab, and also Further analyses tab.

The further analyses tab provide extra analysis features such as enabling Forecast, which runs a time series forcasting algorithm, and plot the forecast as the orange line as in below.