Press "Enter" to skip to content

Category: Visualization

Smoothed Lines and Data Visualization

Published 2026-04-02 by Kevin Feasel

Kerry Kolosko digs into data visualization theory:

Power BI development is a relatively straight forward process when managed by one individual start to finish. But when the development process is shared among team members, ways of working need to be established and common work management frameworks such as agile, lean, HCD and UI/UX Design are adopted.

These frameworks can be useful for teams but as always, the rigid adoption and adherence to frameworks can cause project inefficiencies. It took a fair bit of corporate learning to acknowledge that applying Agile methodologies to a construction project and waterfall methodologies to a software project, weren’t effective.

There’s a lot in here around pros and cons of various tooling (like wireframing), visual selection, the grammar of graphics, and what smoothed lines actually represent. Smoothed lines is a bit of a hobby horse for me, as those smoothed lines represent a model of the data rather than the actual data, so if you show me the former, you’d better also show the latter.

Leave a Comment

Compressing Images in R

Published 2026-04-01 by Kevin Feasel

Yihui Xie announces a new package:

Last month, @bastistician opened an issue on the litedown repo pointing out that knitr has a hook_pngquant() function for compressing PNG plots from code chunks, but litedown lacks such a feature. He included a reasonable workaround—calling system2("pngquant", ...) with litedown::get_context("plot_files") in a chunk at the end of the vignette. It shrank his vignette from 80 KB to 54 KB, which is a 33% reduction. Not bad.

The catch, of course, is that it requires pngquant to be installed on the system. For R users, installing a system binary is more friction than it sounds: it is brew install pngquant on macOS, a separate package manager invocation on Linux, and hunting down a standalone executable on Windows. If you maintain a package that others will build, you are now asking all of them to do this—for every machine they use. By contrast, install.packages("tinyimg") works the same way everywhere, which is the kind of simplicity that makes a tool actually get used.

This is why I created tinyimg.

Read on for more details about how tinyimg works, how well it compresses, and how it integrates with litedown.

Leave a Comment

Maps in Microsoft Fabric now GA

Published 2026-03-23 by Kevin Feasel

Johannes Kebeck makes an announcement:

When we envisioned Maps in Microsoft Fabric, our goal was to empower any data citizen to analyze data in time and space without any specialized knowledge. Introduced in preview at FabCon Europe 2025, it has since been used by customers across industries creating and sharing map-centric applications. Additional features were added at Ignite 2025, and this week at FabCon Atlanta, Maps in Microsoft Fabric is generally available – along with new capabilities that expand how geospatial data can be modeled, visualized, and operationalized at any scale.

Read on to see what’s new in maps.

Comments closed

Prevent Future Date Spillage in Power BI Visuals

Published 2026-03-20 by Kevin Feasel

Kenneth Omorodion lives in the now:

For Power BI developers, one very common (and frustrating) issue is when measures spill into future dates on charts especially when working with some time intelligence DAX calculations (e.g. MTD, YTD, etc.), date dimensions that extend beyond current date, and forecast-enabled tables.

In Power BI charts (e.g. line or bar charts), apart from dates with data, measures are also evaluated for every date on the axis, regardless if there is data or not. For example, if my dates table runs to 2026 December, but my data table only have data up to today, when I create a measure that leverages MTD or YTD for example, Power BI will tend to evaluate the measure for all dates that exist in my Dates table, unless I explicitly apply a logic to prevent this behaviour. This behaviour might result in flat lines on charts, misleading trends, and confusion to intended users.

In this article, I will demonstrate some examples of approaches to prevent or manage future dates spillage in Power BI.

Click through for some tips.

Comments closed

A Sparkline-Enabled KPI Card for Power BI

Published 2026-03-18 by Kevin Feasel

Elena Drakulevska shares a Power BI custom visual:

Sometimes you start experimenting with something small… and suddenly a whole little universe appears.

This happened while I was playing with the idea of a custom KPI card visual in Power BI.

I absolutely love KPI cards, but I’ve never shipped a custom SVG KPI to clients before. Once you do that, they can get a bit… stuck with it.

So I decided to explore a different path.

Following the fantastic tutorial by Phil Seamark, I built my first custom visual!

Click through for the results.

Comments closed

The Challenge of using Questions as Slide Titles

Published 2026-03-05 by Kevin Feasel

Simon Rowe explains a challenge:

The importance of an effective slide title cannot be overstated. Positioned in prime real estate at the top of the page, it is often where an audience’s eyes will land first. With that in mind, it is worth investing time to craft a title that introduces the content below and establishes a clear purpose. Too often, this valuable space is used for purely descriptive statements. Let’s look at an example.

Read on to see one example, showing how the change of titles and a bit of thought around the use of color as an identifying feature can make a big difference for viewers.

Comments closed

Building a Graph for Its Takeaway

Published 2026-02-20 by Kevin Feasel

Cole Nussbaumer Knaflic reminds us that visuals should have purpose:

I was facilitating a workshop recently when someone asked one of my favorite questions about a graph on the screen: “So… what are we supposed to take away from this?”

Such a simple—and useful—question.

One challenge was that the graph was attempting to show multiple comparisons at once, so it wasn’t clear what mattered most. To further complicate things, the data in question spanned very different magnitudes, with one category dwarfing the rest.

Click through for a demonstration and how changing the visual layout can affect the message. The challenge I tend to run into is that, when I’m developing a visual for an application or a report, I don’t know what the precise message should be at that moment in time. I have to design with an idea of the data, but what actually emerges will depend upon what data is in there. Tailoring a visual for a specific message at a specific point in time is a lot easier when building a presentation, but it gets tricky when you’re building an application for the long haul.

Comments closed

Two Options for Content Layout in Power BI

Published 2026-02-16 by Kevin Feasel

Valerie Junk covers a pair of options:

In this tutorial, I want to show a small but very practical formatting setting in Power BI.

When we create a table or matrix visual, we sometimes end up with white space on the right side. For example, if you show data by month and you only have 6 months of data so far, but you design the visual to fit 12 months, the table/matrix is already sized for the full year, which leads to a lot of empty space.
In Power BI we have two column header formatting options:

Click through for the two options, where you can find the option, and some important information around both options.

Comments closed

Choosing between PCA and t-SNE

Published 2026-02-13 by Kevin Feasel

Shittu Olumide visualizes some data:

For data scientists, working with high-dimensional data is part of daily life. From customer features in analytics to pixel values in images and word vectors in NLP, datasets often contain hundreds and thousands of variables. Visualizing such complex data is difficult.

That’s where dimensionality reduction techniques come in. Two of the most widely used methods are Principal Component Analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (t-SNE). While both reduce dimensions, they serve very different goals.

The thing that ultimately soured me on t-SNE is the stochastic nature. You can run the same set of operations multiple times and get significantly different results. It’s really easy to use and the output graphs are really pretty, but if you can’t trust the outputs to be at least somewhat stable, there’s a hard limit to its value.

Comments closed