1
Step 1: Choose Your Input Mode
Pick Path mode if your data has natural hierarchy levels (Region → Category → Product). Use Graph mode for Source + Target edge lists. Hybrid combines both — start simple and upgrade later.
ProSankey Learning Center
Power BI Native Workflow
ProSankey Setup Guide
From your first flow diagram to advanced Pro features — everything you need to get ProSankey working exactly how your data demands.
Guided Learning Path
Follow this sequence for a clean first setup and to avoid the most common mistakes.
2
Step 2: Bind Fields & Check Data Types
In Path mode, drag categorical columns into Levels and a numeric measure into Value. In Graph mode, bind Source Node, Target Node, and Value. Ensure Value is numeric — text fields will produce an empty diagram.
3
Step 3: Review Null & Zero Handling
Rows where Source equals Target, Value is zero, or any required field is null are silently dropped. Check your data model for these cases early — they are the most common cause of missing nodes.
4
Step 4: Configure Colors & Labels
Set per-level colors in Nodes → Level Colors for Path mode. Add Color Rules for condition-based highlighting. Enable Flow Labels and Node Labels (Pro) to display values on the diagram.
5
Step 5: Tune Performance
Enable the cache in Performance settings to skip redundant parse/layout passes on format-only updates. Set Max Nodes and Max Edges if your graph is dense. Viewport culling is on by default and skips off-screen elements automatically.
Action Playbooks
Practical step sequences for common ProSankey tasks.
Action Playbook
First Sankey in 5 Minutes
Zero to a working flow diagram in Power BI Desktop
- 1 Import ProSankey from AppSource and add it to your report canvas.
- 2 Bind 2 categorical columns to Levels and a numeric measure to Value.
- 3 Resize the visual to fill at least 60% of the report canvas for readable node spacing.
Action Playbook
Highlight What Matters
Use color rules to surface key flows instantly
- 1 Go to Format panel → Color Rules. Add a rule with the node name you want to highlight.
- 2 Pick a high-contrast color (red, amber, green) that stands out from the default palette.
- 3 If you have multiple priorities, use rules 1–3 for critical, caution, and healthy states.
Action Playbook
Handle Large Graphs Gracefully
Keep dense diagrams readable and fast
- 1 Enable bucketing (Pro) with a 2–5% threshold to collapse long-tail nodes into "Other".
- 2 Set Max nodes (Performance) to 50–100 to cap initial render density.
- 3 Enable the performance cache to avoid re-running layout on format-only updates.
Feature Reference
Every ProSankey feature explained — what it does, how to configure it, and which plan is required.
Core (Free) Pro Pro display only
Core Path Mode ⌄
Levels data role
Drop up to 3 columns (Free) or 16 columns (Pro). Each column becomes one vertical stage in the Sankey.
Automatic layout
ProSankey computes node positions and band widths from the Value column using the d3-sankey algorithm. No manual positioning required.
Node identity
By default, nodes with the same name in different levels are treated as separate nodes (Per-level). Switch to Cross-level (Pro) to merge same-name nodes across levels.
Null strategy
Configure how null values in level columns are handled: skip the row, treat null as a literal node, or propagate to the next non-null level.
Core Graph Mode ⌄
Source Node & Target Node roles
Each row defines one directed edge from Source to Target with the given Value. Nodes are inferred automatically from the union of all source and target values.
Arbitrary topology
Unlike Path mode, Graph mode supports any directed graph structure — multi-source, multi-target, branching, merging, and crossing flows.
Self-loops
Edges where Source equals Target are treated as cycles and rendered as self-loop arcs if loop display is enabled (Pro), or dropped otherwise.
Pro Hybrid Mode ⌄
Combined binding
Bind both Levels and Source/Target roles simultaneously. The adapters merge both data sources into a single flow graph.
Use case
Ideal when you have a structured hierarchy (from levels) but also need to overlay specific ad-hoc connections that do not fit the level structure.
Pro Flow Labels ⌄
Placement
Center (default), near source, or near target. Labels are positioned along the link band.
Content mode
Show raw value, percentage of grand total, or percentage of source node total.
Min width threshold
Labels are hidden automatically when the link band is narrower than the configured minimum to avoid clutter.
Font & color
Full control over font family, size, bold, color, and optional halo for readability on complex backgrounds.
Pro Node Labels ⌄
Placement
Inside node, outside left, outside right, or outside (auto-sides based on position in diagram).
Content
Show node name, value, percentage — or all three stacked.
Halo
Optional text halo for legibility on colored node rectangles.
Pro display Loop Detection & Cycle Arcs ⌄
Detection
ProSankey runs cycle detection on every update. Circular flows (A → B → C → A) are identified and removed from the d3-sankey layout graph to prevent crashes.
Cycle arcs
Detected cycles are rendered as curved back-arcs outside the main diagram area (Pro). Style the arc color, dash pattern (solid/dashed/dotted), and indicator dots.
Self-loops
Self-referencing edges (A → A) render as small loop arcs attached to the node itself.
Bidirectional flows
Two opposite-direction edges between the same pair of nodes are given separate lane offsets so they are visually distinguishable.
Pro Path Tracing ⌄
Activation
Click any node or link to activate path trace mode. All flows not connected to the selected element are dimmed to the configured fade opacity.
Fade opacity
Configure how opaque non-highlighted elements remain (5%–50%). Lower values give more dramatic highlighting.
Highlight accent
Selected paths use the accent color from your Power BI theme or a custom override color.
Deactivation
Click the selected element again, or click anywhere outside nodes and links, to return to the full-opacity view.
Pro Top-N Bucketing ⌄
Threshold
Nodes contributing less than N% of the grand total are collapsed into a single "Other" node per level. Configurable from 0.1% to 20%.
Bucket label
Customize the display name of the "Other" aggregate node (default: "Other").
Expand inline
Click the "Other" node in the visual to expand it and reveal the collapsed members without leaving the report.
Color
Configure a distinct fill color for the "Other" node to distinguish it from real data nodes.
Core Color Rules ⌄
Rule slots
Up to 8 rule slots, each with a Rule Value (text match on node label) and a color. Rules are evaluated in order; first match wins.
Case sensitivity
Matching is case-sensitive by default. Enable Case-insensitive matching in Data settings (Pro) to ignore casing.
Priority
Color rules take priority over per-level colors and node category styles, but are overridden by explicit node category color assignments.