The Ultimate Guide to Chart Types in Power BI

The Ultimate Guide to Chart Types in Power BI

Data Visualization focuses on presenting data in charts and graphs. The challenge is to get the art right without getting the science wrong, and vice versa.
Data Storytelling involves structuring data into a meaningful narrative that guides toward a conclusion.

Elements of a good data story.

1. Data – Well prepared, accurate, and relevant information.
2. Visuals- Charts and graphs that enhance dashboards.
3. Narrative – A compelling storyline that connects insights with business impact.

Data preparation and modeling in Power BI

1. Import the data in Power BI – CSV, excel, SQL, web
2. Data Cleaning and Transformation
   -Power Query for data cleaning
   -Handle missing values, duplicates, and errors – before visualization.
3. Build a data model.
   -Schema – Star schema -is simpler and faster.
   -Snowflake schema – Define relationships between tables -one to many, many to many, one to one.
4. DAX for storytelling.
   -Use measure and calculated columns -use this properly
   -Common dax formulas -Sum(),Average(),Calculate()

Designing effective Power BI Dashboards.

Dashboards vs Report.
a. Dashboard – interactive visualization that provides a high-
level view of key metrics and insights.
b. Report – a report designed to be printed or exported with a
fixed layout, unlike interactive dashboards.

Principles of Dashboard Design.

Below are the principles for a good dashboard.

1. Know your audience – always ask yourself this question before designing your dashboard. > Who is going to consume the dashboard?
2. Think About The Flow of Data And Power BI Dashboard Layout.
3. Which graph or KPI should I represent first?
4. Is the dashboard too cluttered?
5. How do I prioritize information?
6. How can users drill down into specific data points?
7. What level of interaction provides the most value?
   3.
   a. Clarity & Simplicity – Avoid Clutter.
   b. Visual Hierarchy – Also prioritize important information.
   c. Use of colors & themes – Keep it professional and readable.

Visualization is worth billions or trillions of data items.

Why use data visualizations?

1. Visualizations help us understand complex data.
   The best reason to use a visualization to understand your data is that most data sets are far too large to consume in their raw format. Humans are limited in what information we process and compare in our heads, especially if that information resides in a million-row data set, since human beings are good at processing visual information.
2. A data visualization has to be accurate thus conveying the data. It must not mislead or distort.
   Some pre-attentive attributes are better at showing quantitative (Measured) data and some are better at showing qualitative (categorical) data.

Chart types and their uses.

1. Charts that show trends

a. Line charts.

• They are used to visualize a measure over time to see how that measure changes from point to point.
  line chart views trends in data over time, years, days, months, quarters, and downtrends. Use case
• stock price change over five years
• website page views during a month.

b.Multi-line chart.

• Capture multiple numeric variables over time. It can include multiple axes allowing comparison of different units. The secondary axis is often used in Power BI.
  Use case.
  • Apple vs Amazon stocks over time.

c.Area Chart.
Just like line charts, area charts are used to visualize trends over time. Area charts, allow us to visualize comparisons across multiple categories showing how each category contributes to the overall measure. They do this by filling in each area of the chart with their unique color as demonstrated in the visual on the left.
Area chart – Use for rates rather than totals. Use sensible base geography
use case
-internet usage rates in certain geographies,
-house prices in different neighborhoods.

d.Stacked area chart.

Mostly used for variation of area charts. This is best used to track the breakdown of a numeric value by subgroups.
Multiple data series are layered on top of one another, which is useful for understanding part-to-whole relationships.

2. Part-to-whole.

It means _sehemu hadi jumla _ in swahili. A comparison between a part and a whole of one thing and how it is similar to a part and a whole of another thing.

a). Pie chart & Donut chart.

(i). Pie chart – A circular chart that could present values of a dataset in the form of circle slices.
(ii). Donut charts are a way in which to represent how data is spread across different categorical values.
Donut charts are much the same as pie charts but with a hole in the middle.

b. Treemaps.
These are 2d rectangles whose size is proportional to the value being measured and can be used to display hierarchically structured data.
Method for displaying hierarchical data using nested figures ,usually rectangles.
They require one or more dimensions and one or two measures to build them. Treemap – Show hierarchical data as a proportion of a whole.
Storage usage across computer machines comparing fiscal budgets between years.

Example.
How can you show the part-to-whole relationship between marital status vs Distance from home in (km)

1. Drag the sum distance in (km) to the size mark card and marital status to the color
2. To Distinguish between the different squares, add labels . Drag sum distance in (km) and marital status to the label shelf. The order you drag the fields onto the marks card depends on whether you want sum distance in (km) or marital status to appear on the label first.

Moreso u can edit the labels manually.

1. Click labels on the Tableau shelf and the following window will be displayed.
2. Select the three dots next to the Text option,you can specifically edit what appears on the label.

More notes on Treemaps

A treemap can be used to display part-to-whole or hierarchical relationships using rectangles or squares as the area of the measure.
However, the user could take longer to interpret this chart type over a bar chart, because the human eye can interpret the pre-attentive attribute of length better than the size attribute of an area.
A treemap orders the segments based on highest to lowest with the largest/highest sections being in the top-right area.

To show how the subcategories fit into these categories, all we need to do is click the plus sign on Marital Status or sort them which will drill down to Sub-Category.

The advantage of using treemaps is you can use multiple measures.
I don’t recommend sizing by a measure that has negative numbers,
because to show the subcategory, you would have to use absolute
values, which could be misleading.

c. Heat Map
Show the relationship between two factors. Segment analysis of target market or sales leads by individual rep.
Calender Heatmap.
Question.
Display a single month of a calendar to show the number of orders shipped by day.
Solution
To start with, filter to a single month by dragging the hire Date to filters, selecting Month/Year, and choosing your month.

A calendar heatmap shows you hot and cold spots in a monthly view.

d.Decomposition trees.
The decomposition tree visual in Power BI lets you visualize data across multiple dimensions.They are easy to read and explain.
They are fantastic way to dig into drivers of a specific metric.
It is valuable for ad hoc exploration ,conducting root cause analysis and identifying influential variables which explain the variation in the target variable.
The architecture looks like branches of a tree, hence decomposition TREE.

It requires two types of inputs .

• Analyze – the metric you would like to analyze. It must be a measure or an aggregate.
• Explain By – one or more dimensions you would like to drill down into.
  Decomposition trees are a great method to analyze a target variable.They can be complimented with other visuals or information to build further context (e.g cards ).

Like for instance you can get the total average working years of women in sales executive department who are mid in their career stage and job level 3. Below is the illustration,
so the card shows the average working years.

3.Visualize a single value

a. A card – is great for showing and tracking Key Performance Indicators (KPIs) in dashboards or presentations.
b. Table chart – best to be used on small datasets, usually displays tabular data in a table.

c.Gauge chart -Gauge chart looks similar to a speedometer of a Car. used to measure and gives a visual display of the amount,level or contents of something. Often used in executive dashboards to show relevant KPIs.
A gauge chart is a half-circle chart, which tells whether a target is achieved or not. Gauge charts are generally used to measure the progress toward a goal.

4.Capture Distributions

Distributions – describes how values are distributed for a field,simply terms which values are common and which are not.

a.Histograms
Unlike a bar chart, it shows the distribution and frequency of data converts numerical data into bins as columns.
You can change the bin size in the histogram. This type of chart shows the distribution of your data using equal bin sizes. The bin
sizes can be static, as the solution provided, or you can use a parameter to let the user choose the bin size.
Bins/class each bar height indicates the frequency of data points with a value within the corresponding bin.

b.Box and Whisker
They Show the distribution of a set of the data,
Examples Understanding your data at a glance seeing how data is skewed towards one end, and identifying outliers in your data.

*c.Violin plot *. shows full distributions of the data alongside summary statistics.

5.Relationships.

*a. Scatterplots *
used to show the relationship between two things to understand if a relationship exists.
Scatterplot – Investigate relationships between quantitative values examples Male versus female likelihood of having lung cancer at different ages or technology early adopters and laggards purchase patterns of smartphones.
Scatter plots are my favorite chart type because of their ability to reveal and communicate correlations.

b.Bar Charts.
Bar charts are one of the most popular and effective charts you can use. They allow us to compare measures over different categorical data values easily.
It allows the human eye to compare differences between length or height, one of the preattentive attributes.
There are two types of basic bar charts:
1. Horizontal.
2. Vertical.
Bar charts have other variations, including stacked bar charts, diverging bar charts, and histograms.
Drawbacks of using a vertical bar chart – the reader would have to tilt their head to read the headers.
The benefit of using a Horizontal bar chart – it allows the sub-category header to have more space. It allows the reader not to shift position to read therefore allowing faster interpretation of the chart.
Bar charts are great for comparing whether a subcategory is above or below a constant line.

A constant line is used for charts when you have a specific target or reference point.
Bar – Compare data across categories. eg volume of shirts in different sizes, or percent of spending by department.
a.The length of the bars tells you which category has the
highest value.
b.The axis of the bar chart indicates what that value
represents.

Stacked Bar Chart.
Stacked bar charts are useful for showing potential relationships in data and display an extra level of detail. To enhance this chart type you can add labels to each gender and a total per subcategory.

100 % Stacked Bar Chart(Percent of Total).
Stacked bar charts are useful for showing the contributions of each dimension of data, which is known as part-to-whole relationships. The drawback, however, is that 100% stacked bar charts don’t allow for a comparison across the subcategories for each gender.

100% Stacked bar charts are ideal for comparing segments within each bar. In our example the Percent of Total allows the user to see the percentage of each gender regardless of the total value for that subcategory.
When you see an upward triangle on a green field that means the field has been changed to a Table Calculation.

Shared Axis
You might have a data set that has multiple measures in different columns. A shared axis will allow you to compare two or more measures on one axis. This is especially beneficial if the measures are on a similar scale.

Maps

Maps are a good way to visualize location-based data. Tableau recognizes geographical fields such as country and city and automatically generates latitudes and longitude. This prevents the user from manually looking up coordinates or paying extra for this type of data.
Maps provide an effective and intuitive way to represent geographical data.

Density Map.
Density maps, visualize with intensity where there are larger clusters of data points to make it obvious where the most activity is taking place. An important feature of density maps isn’t just the areas of large clusters but also outliers.
The Density mark shows where the highest levels of concentration are within your scatter plot

*Symbol Maps *

• A symbol map allows you to show multiple measures on a single map.
  Symbol Maps are good for showing exact locations.
• Remember if you are visualizing raw numbers(for example, infection rates), it is important that you use a rate that is relative to the population of the area. This is to allow a baseline comparison.
  Use for totals rather than rates. Be careful as small differences will be hard to see. for instance Number of customers in different geographies.
• A symbol map is a visualization that represents data on a map using a symbol, a symbol being any shape or image we wish to represent the values.
  
  Symbol Map.

Note
When using filled maps, you should use them with caution. Since our eyes are naturally drawn to larger areas on the map.Therefore, when you have smaller countries or states that might have a higher value, their smaller size means you might not instantly see that. This is where a symbol map is better at visualizing the smaller areas.

Highlight table.
The Highlight table is a very easy and effective visualization to build. A highlight table uses diverging colors to demonstrate high and low numerical values across dimensions.
Highlight Table – shows detailed information on heat maps, eg the percent of a market for different segments or sales numbers in a region.

Gantt Chart.

Gant charts provide a visual representation of schedules, tasks, and timelines which can be effective in project management and planning.

• Gantt – shows duration over time eg project timeline duration of a machine’s use availability of players on a team.
  It enables users to track progress, identify dependencies, and monitor the length of activities.
  calculate the difference in date using the function in tableau known as DATEDIFF

Barcode charts.

Gantt has many use cases, and one of those is to build a barcode chart. Barcodes are used to see data and can be read by a machine.

SoundWave.

Bar charts are heavily used in data analytics. The soundwave chart is a version of the bar chart. It uses the standard bar chart approach. The soundwave chart can be used to enhance the visualization of the volume. Soundwave charts are great for showing volume. This type of chart doubles the extreme values.

March 9, 2025 at 01:24PM
https://dev.to/gateru/the-ultimate-guide-to-chart-types-in-power-bi-ii6
Kaira Kelvin.