jkunst.com

What is Joshua Kunst?

Joshua Kunst is the personal site of a data practitioner who builds applications, statistical models and data visualizations. The site presents him as someone who works across the whole path from raw data to something a person can actually look at and use, rather than specializing in only one of those steps.

What the site covers

The page lists several strands of work:

• CensoLab – an application-oriented project, suggesting work with census-style or population data.
• highcharter – a known R package for interactive charts, which places him in the R and data-visualization ecosystem.
• Visual Data Lab and Visual Data Sketches – exploratory visualization work, where the goal is often to make an idea visible rather than to ship a finished product.

The framing is telling: he describes exploring data "sometimes for work, often for the challenge and the satisfaction of making an idea visible." That points to a portfolio built around curiosity and craft, not a single commercial product.

Who it is useful for

If you work with data in R, the highcharter connection is the most immediately practical entry point. If you are looking for examples of how to turn a dataset into an interactive application, the project list is a place to browse. If you are simply trying to understand what this person does before collaborating or hiring, the site functions as a compact portfolio rather than a blog or a company page.

For a concrete scenario: an analyst who needs to build an interactive dashboard in R could look at his charting work for patterns, while someone commissioning a data story might scan the visualization projects to judge style and range.

A practical next step

Start with the project that matches your stack or your question. If you use R, look at the highcharter work first; if you care about exploratory or public-interest data, start with CensoLab or the visual sketches. The "More projects" link is the natural way to see the full range before deciding whether his approach fits your needs.

What data visualization projects has Joshua Kunst created?

Joshua Kunst’s site presents him as someone who builds applications, models and data visualizations, with selected work rather than a full portfolio archive. The projects named on the homepage are:

• CensoLab — a data application or lab-style project.
• highcharter — an R package for Highcharts-style interactive charts, widely used in R workflows.
• Visual Data Lab — a visualization-focused project or application.
• Visual Data Sketches — exploratory visualization sketches.
• More projects — a link to additional work beyond these featured items.

For a practical next step, start with highcharter if you need interactive charts in R, and browse Visual Data Sketches if you want to see his exploratory range. If you are evaluating his applied work, CensoLab and Visual Data Lab are the more application-oriented entries to inspect first.

For related official references, see Joshua Kunst and highcharter.

How can I use highcharter for interactive charts in R?

highcharter is an R package that wraps the Highcharts JavaScript charting library, letting you build interactive charts from R data without writing JavaScript. You pass it a data frame or vector, choose a chart type, and get a chart with tooltips, zooming, legend toggling and hover behavior. It is listed among the projects on Joshua Kunst.

Typical workflow

1. Prepare data in R as a data frame, tibble, list or time series.
2. Call a chart function such as hchart() for a quick chart, or highchart() when you want to add series and options step by step.
3. Layer on behavior with hc_tooltip(), hc_legend(), hc_xAxis(), hc_add_series() and similar helpers.
4. Render it in an R Markdown or Quarto document, a Shiny app, or save it as a standalone HTML file to share.

library(highcharter)

hchart(mtcars, "scatter", hcaes(x = wt, y = mpg, group = cyl)) |>
  hc_title(text = "Weight vs. fuel economy") |>
  hc_tooltip(pointFormat = "{point.y:.1f} mpg")

Where it fits

• Exploratory analysis: hover tooltips and zoom make it easier to inspect outliers and dense points than a static plot.
• Shiny dashboards: charts can react to inputs, and clicking or selecting points can feed values back into the app.
• Reports and teaching: an HTML file or embedded chart travels well and lets readers explore without running R.

Trade-offs to weigh

• It depends on a JavaScript library, so very large datasets can become slow; aggregate or sample first.
• Theming and deep customization mean learning Highcharts-style options, which is more involved than base R plotting.
• Static export for print is less direct than with ggplot2; if the final output is a PDF or journal figure, a static tool is often simpler.

Choosing

Use highcharter when interactivity, tooltips and browser-based sharing are the point. Stay with ggplot2 when you need print-quality static graphics and a large extension ecosystem, and use plotly if you want a similar interactive result with a different API. A practical next step is to rebuild one chart you already have in ggplot2 as an hchart() and compare how much configuration each needs for the interactivity you actually want.

What services does CensoLab offer for data analysis and modeling?

CensoLab is listed on jkunst.com as one of Joshua Kunst's projects, but the site does not detail specific service packages, pricing, or engagement terms. What it does show is the style of work behind the project: turning data into models, applications, and visualizations, with a clear emphasis on making an idea visible rather than only reporting numbers. That suggests CensoLab is best understood as a data-analysis and modeling practice, not a product with published service tiers.

What you can reasonably expect from this kind of practice

• Exploratory analysis – finding patterns and structure in a dataset before choosing a modeling approach.
• Modeling – building statistical or machine-learning models suited to the question, then checking whether they hold up.
• Visualization and applications – communicating results through charts, dashboards, or small interactive tools rather than static reports.
• Prototyping – testing an idea quickly to see whether it is worth developing further.

How to decide whether it fits your project

The site's framing is useful as a filter. If your need is a one-off chart or a standard monthly report, a general analytics provider may be cheaper. If you have an ambiguous dataset, a question that needs modeling rather than summarizing, or a result that must persuade a non-technical audience, the combination of modeling plus visualization is the relevant strength.

A concrete scenario: a team has three years of operational data and suspects that a handful of variables drive most of the variation, but nobody can agree on which. The useful first step is an exploratory pass that produces a small set of clear visuals and a candidate model, which the team can then discuss with evidence in hand.

Next step

Because service details are not published on the site, contact Joshua Kunst directly with a short brief: the data you hold, the decision you need to support, and the audience for the output. Ask specifically whether the work would be scoped as analysis, a model, a visualization, or an application. For related open-source work in interactive charting, see Joshua Kunst and the highcharter project referenced there.

What are Visual Data Sketches and how do they demonstrate data exploration?

Visual Data Sketches are one of the project areas on Joshua Kunst's site, listed alongside CensoLab, highcharter and Visual Data Lab. The page describes them as part of a practice of "finding ways to see data through models, applications and visualizations" — work done "sometimes for work" and "often for the challenge and the satisfaction of making an idea visible." The site does not define the sketches in detail, so treat them as exploratory visual experiments rather than a finished product or a documented tool.

What they demonstrate about data exploration

A sketch, in this context, is a quick visual attempt: a way to try a chart form, an encoding or an idea before committing to a full application. That is the opposite of a polished dashboard. It shows exploration as a process of asking "can this idea be seen?" rather than "does this report ship?"

• Small scope: one idea per sketch, so the visual question stays clear.
• Fast iteration: sketches are cheap to make and easy to discard, which encourages trying several encodings of the same data.
• Visible reasoning: the output is the argument — if the chart does not communicate the pattern, the idea is not yet visible.
• Range over depth: a portfolio of sketches shows breadth of chart types and data situations rather than one deep product.

How to use this as a model

If you are building your own exploration practice, borrow the sketch mindset: keep a folder of throwaway visuals, each answering one question, and only promote the ones that survive a second look into a real application. The distinction is useful when deciding where to spend effort.

For related work in the same space, see Joshua Kunst and the interactive charting library highcharter.

How can I hire or collaborate with Joshua Kunst on a data visualization project?

The site itself does not publish a hiring page, contact form, or collaboration terms, so the practical route is to reach out through the channels listed on Joshua Kunst and lead with a concrete project rather than a general inquiry. The page describes him as someone who builds applications, models, and data visualizations, and who works on data problems both professionally and out of interest in the challenge — which suggests project-based collaboration is the natural fit.

What to put in a first message

• The dataset or domain: what the data covers, roughly how large it is, and whether it is already clean and accessible.
• The decision the visualization should support: exploration, a public-facing story, an internal dashboard, or a model diagnostic.
• Deliverable and format: a static graphic, an interactive web app, a reproducible notebook, or a reusable charting component.
• Timeline and your role: who provides the data, who reviews, and when it needs to be done.
• Budget range or engagement type: paid contract, consulting, or an open collaboration.

Choosing the right engagement

A useful next step: before writing, prepare a one-page brief with a sample of the data, two or three reference visuals you admire, and the single question the audience should be able to answer. That turns a vague “can we work together?” into a project someone can actually estimate. If you are hiring on behalf of an organization, also confirm up front who owns the resulting code and graphics, and whether the work can be shown publicly — the site's own framing suggests a genuine interest in making ideas visible, so public-facing work is likely welcome, but that is an assumption worth confirming rather than relying on.