Communities
Explain participation and community connections with accessible visuals. Personal data stays in the app that owns it.
Try a fictional community report →INTERACTIVE FORCE-DIRECTED NETWORK
Follow the connections.
Select an information node to see what it connects.
Drag, search or switch views to follow the detail.
Try a connectionIllustrative data · hub names shown; select a topic for its full name
Connected informationFictional example
More
Model filter ·
Drag to move · Scroll to zoom · Select a node to explore
No node selected.
Open 3D to load the local renderer. WebGL2 is required.
Drag to orbit · Pinch / scroll to zoom · Tap a node to select and focus. Camera buttons work with keyboard and touch. White marks the selected node; other colours follow the legend. Node size follows connection count. Depth and distance are layout only, not scores or real-world locations. Layout starts still; motion runs only when you request it.
The same filters, selection, labels, spotlight, paths and relationship visibility apply in 3D. Category boundaries and the left-hand layout controls apply to 2D Network. Saved scenes keep separate 3D positions and camera. Export 3D PNG captures this viewport; SVG uses the 2D Network layout.
How your data is connected
Identity review · shared field values
Shared names or field values can describe different people or things. Review the canonical IDs and supplied properties before deciding identity. This tool does not merge records,change aliases,assert owl:sameAs or infer identity from similar spelling. It reviews current filtered nodes only; Model search does not change this scope.
Only nonblank text and finite numbers are grouped. Numeric 1 and text “1” stay separate. Empty/whitespace strings are excluded in both modes. Optional normalization changes comparison only; original values remain visible. Arrays,objects and booleans are not identity keys here. No fuzzy matching or source verification is performed.
One group and 25 records per page. Display previews may truncate long values; Inspect record and graph JSON retain complete fields. Saved scenes keep this review setup. The underlying graph,filters,layouts and exports remain unchanged.
This is an observed model of the current filtered records, using their category and relationship type fields. It does not infer your database tables, constraints, causality or missing connections. Undirected rows use category order, not a source-to-target claim.
Search and page controls only change this summary. Explore applies one model filter alongside your existing filters and opens Network. Category exploration keeps that category and within-category links. Relationship exploration keeps exactly the listed type, categories and direction, plus the endpoints. Clear model filter restores the previous filter scope; Reset clears all view filters.
Node categories
Relationship groups
Idea informed by PuppyGraph’s graph modeling documentation. Dimau Visuals currently summarizes imported files; external database queries and source-table mapping need a separate integration.
Browse connections as an outline
A browsing tree derived from the current filtered connections, not a supplied taxonomy, family tree or reporting hierarchy. Arrows and weights are ignored when choosing parents. Each node appears once; input order breaks equal-depth ties. Loops, parallel links and cross-links remain in Network, Matrix and data exports.
Select a node or parent relationship to inspect its supplied fields in the shared inspector. Expand/collapse changes only this outline. Fifty rows per page; deep indentation is capped visually and exact depth stays labelled. Saving a scene keeps the preferred root, collapsed branches and page. Data/image exports retain their existing graph/2D scope.
Navigation idea informed by Sarrafzadeh and Lank’s hierarchy/network study. This is an original connection outline, not the study’s document-heading tree or its HKG extraction algorithm.
0 nodes0 relationships0 componentsFull graph
This information network is illustrative. Select a node to inspect its purpose, or import your own records.
Text → graph builds word co-occurrences with inspectable source passages. Insights compares degree and shortest-path bridging; these are structural measurements, not semantic or social influence claims.
Website guardrails: 2,000 nodes / 6,000 relationships / 5 MB per file. Category boundaries use supplied categories. Component overview outlines visible weakly connected components; swimlanes separate supplied categories. Neither represents inferred communities. Matrix shows up to 40 visible nodes. No million-node benchmark is claimed. Saved scenes stay on this device and browser.
YOUR VISUAL EXPLORATION PLATFORM
Discover what connects
your information.
Start with a living network of information. Follow a connection, inspect a node, then bring your own data into the picture.
CSV & JSON imports · Browser-local exploration · No upload required
A different view of the same world 01 / CONNECTED
YOUR DATA. YOUR PERSPECTIVE.
01 Bring it together
02 Follow the relationships
03 Keep the insight
CHARTS · DASHBOARDS · STORIES
Make your data
worth exploring.
Bring a table. Find the right view.
Turn your findings into a report.
Build with your CSV, TSV or row-based JSON. Sixteen working chart types, linked dashboard filters and a narrative report builder extend the graph workspace below. Your imports remain in this browser.
Explore the chart library ↓Data and chart
Scatter uses numeric X and Y. Line/area use numeric or ISO-date X. Histogram uses Y. Timelines use X=start, Y=end. Radar requires a common measurement unit.
The filter updates this chart, dashboard and new story cards together. Saved cards retain their snapshot.
Paste a table
5 MB · 10,000 rows · 100 columns. Excel files can be saved as CSV first. A failed import keeps your previous data.
Explain the finding
Source labels and narrative are supplied by you. The studio computes chart values; it does not generate causal claims.
Explore your table
Inspect data quality and column types
TELL THE STORY
From a chart to a clear report.
Build an ordered sequence of findings. Each card retains its chart, filtered data, explanation and source. Export a self-contained report to read, share as a file, or print to PDF.
Exports contain the included chart values. Share them intentionally. Template links contain no imported data. Local saves are specific to this browser.
0 / 12 story cardsCHOOSE BY YOUR QUESTION
A visual vocabulary.
A practical starting point.
Compare. Explain a change. Show a distribution.
Choose the view that fits the evidence.
The working studio charts are original Dimau Visuals components. Additional techniques link to Data Viz Project for guidance and are explicitly marked reference only. The graph workspace retains its separate layouts and examples.
ONE VISUAL SYSTEM, MANY USES
Useful inside your apps.
Learning
Show progress by subject and make practice patterns visible. Use exact counts and a table alongside every chart.
Try a fictional learning report →Culture and discovery
Help readers explore collections and compare themes. Source every claim and make the next relevant item easy to find.
Try a fictional discovery report →Operations and visibility
Explore visits, signups and completions using a shared filter. Imported analytics remain separate from fictional templates.
Try a fictional engagement report →What is available, and what needs an integration?
Charts, CSV/TSV/JSON imports, calculated summaries, data-quality inspection, filtered dashboards, editable stories, portable reports and template links work here. Existing graph exploration, layouts, 3D network view and saved scenes remain below.
The portable DimauVisuals component accepts row data through an in-app JavaScript API. It makes no network requests and creates no accounts. Live SQL/Sheets/API connectors, hosted report storage, AI/Python execution, arbitrary vector editing, medical illustration libraries and proprietary vendor engines are not installed. Product references describe those capabilities without pretending they are enabled here.
Research behind these capabilities ·24sources
Public feature patterns were inspected with Firecrawl and official GitHub REA. These are original implementations. No vendor runtime or full product clone is claimed. Prices are omitted because plans and terms change.
| Source | Observed pattern | Applied here | Boundary |
|---|---|---|---|
| Data Viz Project | Chart discovery by family, input and purpose | Searchable technique library and 16 working chart types | Additional techniques are reference links, not enabled renderers |
| Information is Beautiful | Editorial data stories and topic discovery | Headlines, explanatory notes, source labels and ordered story cards | No copied artwork, article bodies or proprietary datasets |
| Smartsheet | Project timeline, resource and status messaging on a demo request page | ISO-date project timelines and operational report template | No form submission, trial signup or claim of access to the gated product |
| Estuary | Comparison of open-source analytics and data integration options | Separation of ingestion, calculation, rendering and reporting | No live data pipeline or database connector installed |
| Kent State Libraries | Statistical software learning/resource directory | Explicit calculation definitions, missing-value inspection and portable CSV | No statistical package or executable notebook environment installed |
| AnalyzeData | Upload, inspect computed findings, assemble a report | Table import, computed summaries, chart specifications and report export | Deterministic local calculations; no AI/Python execution or hosted reports |
| GitHub analytics-tools topic | Discoverable public project index | Documented public research provenance and portable component boundary | Topic listing is not evidence that every repository is safe, licensed or integrated |
| SciChart examples | Coordinated dashboard ranges and configurable chart styling; public TypeScript examples | Shared row filters, chart metadata and app-local renderer API | Independent SVG components; no SciChart runtime, licence key or claimed real-time throughput |
| Visme | Business visual creation, charts and branded content workflows | Unified creation workspace and labelled visual reports | No vendor assets, interactive media editor or paid engine |
| Venngage | Infographic/report creation and accessible design offering | Exact-value tables, labels, keyboard-native controls and source descriptions | No claim of automated WCAG certification |
| Piktochart | Guided visual/document creation workflows | Five purpose-led templates and compatible chart suggestion | No generated text claims or document-to-graphic AI service |
| Infogram | Charts, dashboards, reports and sharing | Filtered dashboards and portable visual reports | No live SQL/Sheets connection or hosted collaboration |
| Canva | Template-led graph creation | Immediate examples, data-field selection and exports | No imported templates, stock library or general canvas editor |
| Adobe Illustrator | Custom vector and illustration workflows | Editable SVG chart export | No Illustrator engine or arbitrary vector editing |
| Mind the Graph | Scientific visual communication and illustration offering | Sourced narratives and precise chart descriptions | No medical illustration library or scientific validation claim |
| Gamma | Presentation, document and visual content creation | Ordered report cards with an introduction and title | No Gamma runtime or AI generation |
| Flourish | Interactive charts and data storytelling | Chart-to-story workflow, shared filters and portable report | No Flourish templates, scrollytelling engine or chart animation copied |
| Power BI | Enterprise analytics and reporting offering | Typed data, explicit aggregation and coordinated views | No Power BI/Copilot/data-model engine integrated |
| Tableau | Visual analysis and data application offering | User-selected fields, alternative views and source-aware inspection | User-supplied Ask Data feature wording was not assumed current |
| Looker Studio | The supplied address currently returns Data Studio branding and report/dashboard messaging | Filter-aware reports and dashboards | No Google account connection; current naming observed rather than assumed |
| Yellowfin | Data storytelling with business context | Narrative, source and chart in one report card | No live BI engine or collaborative publishing service |
| Julius | Current homepage presents a general task interface and Excel/model examples | Clear distinction between computed data and authored explanations | No analyst agent installed; generic older feature claims not treated as verified |
| Bricks | Example data visualizations and editable starting points | Working example templates that load immediately | No vendor templates or AI engine copied |
| Zebra BI | Business reporting visuals and Excel/Power BI context | Comparable numeric scales, exact values and structured reports | No proprietary add-in or IBCS compliance claim |
REA static evidence:262nodes/338relationships from acquired public source. It does not establish private product internals or runtime throughput. The supplied Smartsheet URL is a request form;no form was submitted. Additional chart references and the two SciChart demos are retained in the investigation.
ONE DATASET. MORE WAYS TO UNDERSTAND IT.
Borrow the best ideas.
Make them work together.
Follow a connection.
Search properties and tags, focus one or two hops, inspect relationships and trace a shortest path. Add your notes and discover records with shared tags or neighbours.
Explore your graph ↗Change your perspective.
Grouped, force, circle, breadth-first, radial, directed flow, degree lens, component and category-swimlane layouts. An optional 3D orbit view, notes, table, matrix, Data Model and insights share one filtered graph. Quick find and selection history keep you oriented.
Try another view ↗Take your findings with you.
Export JSON, CSV, DOT, SVG or PNG. Save a scene with its graph, notes, filters, positions and camera. Bookmark useful tools in this browser.
Build a scene ↗What needs a separate integration?
Live Neo4j/Cypher or RDF connections, shared accounts and cases, alerts, geospatial analysis, VR/AR, Jupyter execution and proprietary engines are not enabled here. Explore their official products in the library. The existing private AI, Bible and project panels stay in your local saved workspace.
THE GRAPH TOOL LIBRARY
Fifteen approaches.
Find the right fit.
Explore every tool from the reference article.
Select up to three to compare.
Tool names belong to their respective owners. Inclusion is a research reference, not an endorsement or an installed integration. Licensing labels are broad categories; confirm current terms with the vendor.
START WITH YOUR QUESTION
A graph is useful when
relationships are the answer.
People and communities. Dependencies and flows.
Knowledge, research and linked records.
Scale
Measure responsiveness with representative data. Node counts alone do not describe layout or rendering cost.
Output
Choose portable data for analysis, vector images for print, and interactive views for exploration.
Control
A desktop GUI speeds up discovery. A programming library gives more control over a product’s interactions.
Maintenance
Check current documentation, issue activity, licence terms and support before depending on a tool.
Workflow
Keep notebooks, graph databases and browser products in mind. The right integration depends on where the data lives.
Which option fits a large network?
Start with tools designed for your graph’s size and rendering needs, then benchmark real records. Cosmograph, Gephi and Graphviz target different exploration and rendering workflows. Dimau Visuals states its own smaller guardrails above.
How do Gephi and Graphviz differ?
Gephi offers hands-on network exploration. Graphviz lays out text-described graphs for repeatable diagram generation. Dimau Visuals can export DOT for a Graphviz workflow.
What should a developer choose?
Compare the renderer, interaction model and licence. Cytoscape.js, react-force-graph and yFiles for HTML serve different application needs; Graphviz fits automated diagram generation.
What fits research or notebooks?
Gephi and Cytoscape offer desktop research workflows. yFiles Graphs for Jupyter keeps interactive graphs beside notebook analysis. NetworkX node-link JSON can be explored here.
Which formats matter?
Keep JSON for full graph properties, CSV for relationship tables, SVG for scalable images and PNG for quick sharing. DOT supports Graphviz. PDF can be made with the browser’s print dialogue.
Can a browser handle a very large graph?
Rendering technology, layout, labels and graph density all affect performance. This website limits imports to 2,000 nodes and 6,000 relationships; it does not claim GPU or million-node capacity.
Should I use a no-code app or a library?
Use visual exploration when readers need to investigate records directly. Use a library when you need to design the surrounding experience. This workspace provides an immediately usable browser interface.
Is open source suitable for production?
A licence alone does not settle operational readiness. Review maintenance, security, testing, support and data access controls for the deployment you intend to run.
BUILT WITH EVIDENCE
Clear sources.
Honest boundaries.
This is an original Dimau Visuals implementation, informed by public documentation, open-source patterns and REA investigation. It does not reproduce proprietary products or their full feature sets.
Reference article & product distinction
Guideflow’s 15 graph visualization tools ↗
All 15 entries are represented with original summaries and explicit capability coverage. The article’s NeoDash link targets neodash.ai; the graph-dashboard entry here links Neo4j’s NeoDash project and makes that distinction explicit. Neo4j marks its Labs edition unsupported and links a separate supported commercial edition. Article prices are not repeated.
REA and open-source foundations
Official REA ↗ supplies static evidence from the reference page’s public scripts. It does not establish proprietary tool internals. The saved investigation records hashes, evidence and limitations.
D3 force ↗ powers the original layout. NetworkX ↗ powers analysis in the saved local workspace. This website uses independent browser algorithms. Existing licence notices are retained.
How the analysis is defined
Degree counts incident relationships, including multiplicity; a self-loop contributes twice. Components and focus ignore direction. Paths minimise hops, optionally respecting direction. PageRank uses 30 iterations with damping 0.85 and redistributes dangling mass. Density uses unique non-loop undirected pairs. Layout distance has no measurement meaning.
Data privacy & portable formats
This website makes no file uploads or database requests. Imports remain in page memory until you save a scene locally or export a file. Browser storage may be cleared or unavailable. JSON supports nodes plus links/edges, including NetworkX node-link records. CSV imports relationship lists with column mapping; isolated nodes require JSON. The existing local workspace uses your original SQLite database.