iiinova — where your digital self keeps evolving

Daniel Zvinca @danz68.bsky.social · 08/04/2026

The problem is that the damage is caused by only a handful of lunatics, yet everyone ends up paying for it, simply because those responsible are so few. Collective decision control is what is missing. Since ever...

020

Daniel Zvinca @danz68.bsky.social · 02/04/2026

I quickly searched for La Mancha. No, is not close to the border. You need to pick another place. Unless you want to travel, of course.

110

Daniel Zvinca @danz68.bsky.social · 02/04/2026

That kind of “consistency” likely comes from job stability across the entire chain of command: no need to upgrade skills, no fear of replacement, deep resistance to change. It becomes an almost monolithic structure, which also means it will most likely collapse one day, and probably all at once.

110

Daniel Zvinca @danz68.bsky.social · 01/04/2026

A far better spread approximation can be achieved with ellipses. However, although the mass distribution is clearly improved, they still introduce a compromise in the name of geometrical simplicity that does not justify the gain. Still, it is interesting to simulate that as well.

000

Daniel Zvinca @danz68.bsky.social · 01/04/2026

Observable is a generous platform, and for exposure it may be worth it. Beyond that, I am not sure this project would gain much from it. Observable comes with its own constraints, conventions, functional patterns, and computational limitations. (project statistical calculation are done in Python).

000

Daniel Zvinca @danz68.bsky.social · 01/04/2026

The previous screenshot shows how the same mass is actually spread along the wage axis: circle versus elongated violin. My point is that GoG is not universally adequate by default, and this case exposes that weakness rather clearly.

100

Daniel Zvinca @danz68.bsky.social · 01/04/2026

Approximating a mass distribution with circles imposes a pseudo-symmetry and a spread around a chosen anchor (median), unrelated to the actual distribution. As my simulation shows, the circle-based packing model is nowhere near the true statistical model. (pause the animation and use hovering)

110

Daniel Zvinca @danz68.bsky.social · 01/04/2026

Designing interactive visualizations that follow the full depth of data complexity, so that anyone can stop at the level of understanding they feel comfortable with, is nearly impossible, given the way knowledge is actually built and layered. Yet is still nice to try, isn't it?

000

Daniel Zvinca @danz68.bsky.social · 01/04/2026

Jokes aside, I truly believe that any creation, visuals included, sells when it speaks the audience’s language. Using dataviz as a bridge that brings people closer to the true structure and behavior of the data is something else entirely. That is more like chasing the unknown. 1/2

100

Daniel Zvinca @danz68.bsky.social · 01/04/2026

What I worked on most was the Guide. With some help from AI, mainly for grammar correction, I rejected countless versions before its current form, one that mirrors quite closely the way I think. Read it and see for yourself how a geek-like mind looks at things. It may well recalibrate your wishes.

210

Daniel Zvinca @danz68.bsky.social · 01/04/2026

The guide is, in many ways, the post itself. It moves through several perspectives: usage, data, modeling, programming, visuals, and reflection. It is a journey through the thoughts that began the moment I first saw beeswarm encoding and realized it was not accurate in any meaningful sense.

100

Daniel Zvinca @danz68.bsky.social · 01/04/2026

Thanks, Jorge. Having twins may induce a bit of a ... duplication bias, though.

110

Daniel Zvinca @danz68.bsky.social · 01/04/2026

Band size and movement can be adjusted as well (read the guide for interactions). In the global view, try the circle and ellipse simulations. Stop the animation right at the start, then hover and compare. Approximating an occupation with a circle is, statistically speaking, deeply misleading.

100

Daniel Zvinca @danz68.bsky.social · 01/04/2026

Tap, rather than just hover. A tap pins the tooltip so it can be moved and interacted with. Double tap reveals what is going on, whether on a violin or on the axis, each with its own behavior. Right click (or three dots) opens the menu and options.

100

Daniel Zvinca @danz68.bsky.social · 28/03/2026

It also includes a built-in guide, far larger than a normal article, covering concept, math, dataviz, and development.

000

Daniel Zvinca @danz68.bsky.social · 28/03/2026

Desktop or mobile, with rich interactivity, experimental ideas, and a few easter eggs: center in-out size sorting, light border separation, adaptive/pinned tooltip, focus and zooming concept, unusual stack tracking, adaptive scale, shape morphing, and a nod to the law of large numbers.

100

Daniel Zvinca @danz68.bsky.social · 28/03/2026

Sharing the current stage of an interactive violin visualization project. The concept is already there, even if the implementation is not yet as clean as I’d like. iOS rendering quirks forced heavier drawing logic than expected. danz68.github.io/visualprojec...

100

Daniel Zvinca @danz68.bsky.social · 28/03/2026

Desktop or mobile, with rich interactivity and a few easter eggs (unusual stack tracking, adaptive scale concept, shape morphing).

000

Daniel Zvinca @danz68.bsky.social · 20/10/2025

Monotonizing data. Classical PAV: monotone but step-flattened. TPM: monotone and trend-faithful, maintaining readable dynamics such as endpoints, mass, slope continuity, inflection timing, and relative growth/decay patterns.

010

Daniel Zvinca @danz68.bsky.social · 21/06/2025

Numbers and related fields are just a hobby for me, so hearing that any of my posts are helpful to people in dataviz is honestly very flattering.

010

Daniel Zvinca @danz68.bsky.social · 21/06/2025

CSP is always useful for investigating relationships during the exploratory phase of data analysis. The reason it rarely works as an explanatory solution is that we can’t exactly say: "This hard-to-read graph shows there’s no clear relationship using this method..."

000

Daniel Zvinca @danz68.bsky.social · 21/06/2025

Ha, glad to hear the method's been helpful. Measuring or even just describing relationships between variables has been always a tricky business.

000

Daniel Zvinca @danz68.bsky.social · 21/06/2025

Ah, so that’s why my posts/remarks get nearly no traction...🤭

110

Daniel Zvinca @danz68.bsky.social · 27/05/2025

Are you aware of any data visualization designs that use optical illusions to enhance the intended message?

382

Daniel Zvinca @danz68.bsky.social · 06/05/2025

While traditional beeswarms aim to reduce empty space through tight packing, my method focuses on frequency accuracy, with the space being used incidentally rather than as a packing objective.

010

Daniel Zvinca @danz68.bsky.social · 06/05/2025

Here is the result of my NEW frequency dot plot, an arrangement based on data density calculation using the dot size as resolution (granularity). This result actually validates the basic beeswarm packing for this dataset. Beeswarms are often poor density estimators due to their packing artifacts.

110

Daniel Zvinca @danz68.bsky.social · 15/03/2025

It was an exciting experience meeting so many statistical geeks in one place. Meeting @xangregg.bsky.social in person after more than a decade of social media debates was a particular delight. As he said, we could have talked forever.

010

Daniel Zvinca @danz68.bsky.social · 01/03/2025

Not really a CSP. It is a multi line/dot chart. Time and ages are just shifted variables, no chess player trajectory will have curve twists.

010

Daniel Zvinca @danz68.bsky.social · 20/02/2025

When density continuity is a fact my method uses a different idea than KDE. KDE uses a quite large bandwidth (constant or not) to calculate each value "contribution" to the density shape. My method finds the smoothest density shape assuming each value was measured within a given tolerance.

000

Daniel Zvinca @danz68.bsky.social · 20/02/2025

My "best" guess is in the first image (truly, just a guess, no idea how good that is). Then I try to fit the dots as good as I consider is needed, starting from the optimal pack (hexagonal) to no errors at all. The final shape resemblances my "best" guess, an 0.2x error looking like this:

000

Daniel Zvinca @danz68.bsky.social · 20/02/2025

I think that the exploratory phase ends when we are confident enough that we found the density shape (iteratively challenging continuity, gaps, tails). Once we get there, we do our best to fit the available data by minimizing the error placement and/or overlapping, depending on the method.

210

Daniel Zvinca @danz68.bsky.social · 20/02/2025

Using the distribution word in a statistical sense for a few discrete values as those from a Likert scale is a challenge. Can you share the data?

100

Daniel Zvinca @danz68.bsky.social · 20/02/2025

Any chance you can share the data?

100

Daniel Zvinca @danz68.bsky.social · 20/02/2025

Not sure how relevant is the rate. No idea how the system works there, but I doubt they can accommodate as many as they want. I would rather encode how many more applicants than the available places are (assuming all taken) with the following title: More and more people are interested in education.

010

Daniel Zvinca @danz68.bsky.social · 15/02/2025

it looks like your display has a few bugs

010

Daniel Zvinca @danz68.bsky.social · 09/12/2024

Cant really follow these graphs, but I am not familiar with data either. I cant see an explicit functional model to study a relationship from the dots.

000

Daniel Zvinca @danz68.bsky.social · 09/12/2024

I think this example would be a fantastic exercise for people with different backgrounds, but with a bit of functional math and statistical knowledge. I am not sure how a two variable relationship dissertation would look to make it accessible to a broader audience, though.

000

Daniel Zvinca @danz68.bsky.social · 09/12/2024

As you said, I kept staring at it. I love math and the logic behind PCA, but I dont see how it would improve the practical insights for a general audience. PCA might simplify mathematical analysis by removing the correlation, yet it doesnt mean it also simplifies human sensemaking.

200

Daniel Zvinca @danz68.bsky.social · 08/12/2024

What I would conclude: 1. The concentric elliptical isolines might be the representation of a bivariate normal distribution. 2. The nearly constant orientation of the ellipses indicates also a linear relationship where values matter the most, in areas with high density.

100

Daniel Zvinca @danz68.bsky.social · 08/12/2024

It wasn't the case here, because of the orientation, but I didn't know how the isolines shapes look. The ellipsis clearly have an angle. The "rotation" spreads the variance across both axes, so rescaling any axis will not convert the ellipse into a circle.

110

Daniel Zvinca @danz68.bsky.social · 08/12/2024

In a scatterplot, the visual shape of the distribution (ellipse or circle) is influenced by the scaling of the axes. Adjusting the scales can transform a 0 or 90deg oriented ellipse into a circle. This means the elongation might be a plotting artifact rather than a true feature of the data.

120

Daniel Zvinca @danz68.bsky.social · 08/12/2024

I would love to see some isodensity or probability contours on this dataset with adjusted scales based on that. I would not filter out anything, I would just rescale the axis based on the ratio of, say, 90% probability ranges.

110

Daniel Zvinca @danz68.bsky.social · 08/12/2024

I can't really see a functional relationship (not really predictable). However, the visible ellipse rather shows (kind of) a statistical dependency, but only if the range calculated scales dont foul us (if that would be a circle it would be no dependency).

110

Daniel Zvinca @danz68.bsky.social · 23/11/2024

Irrelevant reply: "I am not sure if all need to be data encoding projects. If not: recreate a logo. Describe how much fun you had getting there. Else: recreate a masterpiece. Don't bother admiring too much the original. Convince your audience that your perspective is valid."

110

Daniel Zvinca @danz68.bsky.social · 21/11/2024

... an obvious insensitivity to local variations (if two years in a row health investments were reduced didn't automatically reflect into life expectancy, not those years, nor 10 years later). When such lag and obvious slow data responsiveness occur, CSP is hardly a solution. 🤷‍♂️

000

Daniel Zvinca @danz68.bsky.social · 21/11/2024

Thing is that CSP has often a serious drawback inherited from SP, but rarely mentioned. In the famous CSP showing Life Expectancy vs Health Expenditure a "little" detail was left out. That was the obvious cause-effect lag. For this dataset we can easily guess not only a decade gap, but also ->

100

Daniel Zvinca @danz68.bsky.social · 21/11/2024

To be honest this sort of debate resemblances quite well the survival bias, focusing or counting what is well known ignoring the rest just because they didnt occur too often in practice. Everything I wrote about CSP strengths is legit. It just happens they are not too often reasons for CSP design.

100

Daniel Zvinca @danz68.bsky.social · 21/11/2024

It can be, but how relevant is this? Sure, counting preferences matter, more platforms, even nicer. Yet this doesn't put a value on CSP, but on those poorly designed. I hardly "like" any dataviz technique. I prefer to choose them well when appropriate.

100