Skip to main content
SimilarExport ZIP
100%

ggvis issue - multiple layers on different dataset

✓ Published0🌍 Public
RRCura
Last edited Dec 26, 2014
Created on Dec 26, 2014

This example shows a correspondence analysis of three random normal variables, with points colored by cluster and overlaid with 70% confidence ellipses. It first builds the analysis in R using `ade4`'s `dudi.coa`, then computes clusters via `hclust` and ellipses with the `ellipse` package. The visualization attempts to combine both layers in `ggvis`, using `layer_points` for the observations and `layer_paths` for the ellipses, but reveals a known issue where `group_by` on the path layer is not respected when mixing data frames, leaving the ellipses ungrouped in the final plot.

AI-generated description

Similar vizzes

Loading thumbnail…

Gist demonstrating an R error

This example demonstrates a subtle ggplot2 error that occurs when mapping the `size` aesthetic in a scatter plot while also adding multiple `geom_path` layers for circles. The code uses the `ggplot()` function with `aes(x, y, size = s)` and `geom_point()`, then attempts to overlay concentric circles via `geom_path()` using a helper function `make.circle()`. The error arises only in `make.plot.1()`, which combines both the size mapping and the circle paths, while `make.plot.2()` (no size mapping) and `make.plot.3()` (no circles) run successfully, isolating the conflict between these features.

LLeeMendelowitz
60% match
Loading thumbnail…

The Problem of Occlusion

This example illustrates the problem of occlusion in scatter plots, where overlapping points hide underlying data. The visualization plots automobile displacement against horsepower using data from a CSV file fetched via `d3.csv`, with tooltips displaying vehicle names. Built with the `vega-lite-api`, it renders points as filled, opaque circles, and the dense clustering reveals how marks obscure one another. The code customizes axes and labels through a Vega-Lite config object to improve readability under the heavy overlap.

CCurran Kelleher
58% match
Loading thumbnail…

Gist fe5834a6c79855e23c3c7e3ae7e7f048

This example visualizes the distribution of a dataset through an interactive scatterplot, encoding three variables on the x- and y-axes and point size. The visualization uses D3.js with `d3.forceSimulation` to arrange overlapping points into a readable layout, while `d3.scaleLinear` and `d3.axisBottom` handle axis generation. Rendering is done via SVG circles, and transitions are applied using `transition.attrTween` to smoothly update positions when data changes. The underlying data originates from a static JSON array embedded in the gist, with no external data source, and the code includes hover tooltips via `d3.select().on("mouseover")` for individual point inspection.

AAlexDaGr8
58% match
Loading thumbnail…

Circle-Packing

This example visualizes a hierarchical dataset as a circle-packing layout, where each leaf node is a circle sized by its value and nested groups are represented by larger containing circles. It shows the structure of a module tree from a CSV file, with color indicating depth and text labels clipped to leaf circles. The code uses `d3.stratify` to build the hierarchy from flat data, `d3.pack` for layout, and `d3.scaleSequential` with the `interpolateMagma` color scheme. Hovering highlights a node and all its ancestors via `d3.selectAll` with the `node--hover` class, rendered as SVG elements in a `<g>` container.

HHugoberry
58% match
Loading thumbnail…

Pseudo Cluster Visualisation

This example visualizes a pseudo cluster analysis using an interactive SVG scatterplot. It shows a set of points, colored pink and yellow, distributed across a white canvas, with a large black-outlined circle suggesting a cluster boundary. The rendering relies on raw SVG elements, including `circle` tags with specific coordinates and fill colors, and CSS `:hover` styles that turn points black on mouseover. The visualization is a static, hand-drawn representation of clustered data, using only basic SVG and HTML.

YYiorgos Bagakis
58% match
Loading thumbnail…

D3 ConicEqualArea Projection + D3 Generated Layers

This example demonstrates a multi-layered D3 map visualization using a conic equal-area projection, combining graticules, country choropleths, US county choropleths, geohash cells, heat points, and a graph network overlaid on the same geographic canvas. It uses the HPCC Systems visualization library, with layers such as ChoroplethCountries, GeoHash, ChoroplethUSCounties, Heat, and Graph, all rendered via D3’s geographic path machinery. The data includes country weights, geohash counts, county values, heat coordinates, and graph nodes, each styled with distinct palettes and opacities to reveal spatial patterns. The map is interactive, centered at the equator with a zoom level of two.

GGordonSmith
58% match
Loading thumbnail…

ARL Salaries with ggvis

This interactive scatterplot uses ARL library statistics to examine the relationship between staff counts and total expenditures, with each point representing an institution. A linear regression model with standard error bands is overlaid on the full dataset, and a dropdown menu enables highlighting a selected institution in red. The visualization is built with the R packages ggvis and dplyr, using `layer_points`, `layer_model_predictions`, and `input_select` for the interactive filter, with data loaded from an Excel file via readxl.

CCliffordAnderson
58% match
Loading thumbnail…

Gist fdfc88707da3d0b9b203

This example shows the clustering and connection of points on a map based on their geographic proximity and shared attributes. It uses a force-directed layout via `d3.forceSimulation` to position nodes while links are drawn as curved paths, with a base projection provided by `d3.geoPath` and `d3.geoMercator`. The data is loaded from an external CSV file, and the visualization updates dynamically as the simulation runs, highlighting node neighborhoods and their interconnections.

MMSCAU
57% match