fill3 — Create 3-D filled polygon patches in MATLAB and RunMat.
fill3 creates filled polygon patch objects from X, Y, and Z coordinate data in 3-D axes. It reuses the shared patch pipeline, so handle behavior and patch name-value properties follow MATLAB semantics.
Syntax
h = fill3(X, Y, Z, C)
h = fill3(X1, Y1, Z1, C1, ..., Xn, Yn, Zn, Cn)
h = fill3(X1, Y1, Z1, C1, ..., Name, Value, ...)
h = fill3(ax, X, Y, Z, C)
h = fill3(ax, X1, Y1, Z1, C1, ..., Xn, Yn, Zn, Cn)
h = fill3(ax, X1, Y1, Z1, C1, ..., Name, Value, ...)Inputs
X- Type
NumericArray- Required
- Yes
- Description
- Polygon X coordinates.
Y- Type
NumericArray- Required
- Yes
- Description
- Polygon Y coordinates.
Z- Type
NumericArray- Required
- Yes
- Description
- Polygon Z coordinates.
C- Type
Any- Required
- Yes
- Description
- Color specification.
groups- Type
Any- Required
- Variadic
- Description
- One or more `(X, Y, Z, C)` polygon groups.
props- Type
Any- Required
- Variadic
- Description
- Trailing patch name/value property pairs.
ax- Type
AxesHandle- Required
- Yes
- Description
- Target axes handle.
Returns
h- Type
Any- Description
- Handle or row-vector of handles for created patch objects.
Errors
RunMat:fill3:InvalidArgument- When
- Input groups, axes targeting, or trailing property pairs are invalid.
- Message
fill3: invalid argument
RunMat:fill3:Internal- When
- Internal patch construction or rendering fails unexpectedly.
- Message
fill3: internal operation failed
RunMat:plotting:TooManyOutputs- When
- More than one output is requested from a coordinate chart constructor.
- Message
plotting: too many output arguments
Behavior
fill3(X, Y, Z, C) creates one or more 3-D filled polygons from matching coordinate arrays.
Equal-length row or column vector coordinate inputs create one polygon regardless of orientation.
When any coordinate is a matrix, coordinate vectors matching its row count are shared across its columns; square matrices are deterministically grouped by columns.
Matrix coordinate inputs create one independently addressable Patch object per column and return one handle per column.
Multiple X,Y,Z,C groups in one call create multiple patch objects and return a row vector of patch handles.
A single polygon group returns a scalar patch handle.
fill3 initializes the target axes with the standard 3-D view, so even polygons in the z=0 plane render as 3-D patch content.
Trailing patch property name-value pairs are applied to every polygon group in the call.
The returned patch handles work with get, set, ishandle, and isgraphics.
Color strings and RGB triples are supported through the shared patch color parser.
Numeric C is retained as authoritative CData for indexed color mapping; color strings and RGB triples use the shared Patch color parser.
Integer XData, YData, ZData, and CData preserve their original class, exact values, and shape independently of renderer geometry.
Options
FaceColor accepts color names, RGB triples, flat, or none through patch.
EdgeColor accepts color names, RGB triples, or none.
FaceAlpha and EdgeAlpha accept numeric values from 0 to 1.
LineWidth, DisplayName, and Visible use the shared plotting property model.
Explicit axes handles are accepted as the leading argument, matching other RunMat plotting builtins.
Rendering and residency
fill3 is a plotting sink. It gathers GPU-resident coordinate and CData inputs exactly through their owning providers, then lowers each vector or matrix column to an independent Patch object.
fill3 is a plotting sink and fusion barrier. GPU-resident coordinate inputs are gathered before patch construction because MATLAB-compatible parsing requires shape validation and polygon assembly.
fill3 shares patch rendering behavior, so host and gathered-GPU coordinate inputs produce the same patch objects and graphics handles.
Rendered geometry flows through the shared GPU plotting renderer after the patch objects are built.
A future direct GPU path should preserve the same X,Y,Z,C grouping and patch property semantics while replacing only the coordinate gathering and geometry packing stage.
Declared source precision and class coverage
single
double
Coordinate broadcasting
none
Runtime contract and boundaries
Graphics results are host double handles; fill/fill3 may return a row of handles for independently created polygon columns/groups. Sphere instead supports zero through three requested outputs: statement form plots, while output forms return host double coordinate grids. Handle aliases, value-selected options and exact graphics receiver roles are resolved by Runtime, not guessed from scalar facts.
The existing numerical and geometry owners validate coordinates, integer controls, shapes and properties. Single/double and all eight integer source classes retain their native storage wherever the object has authoritative source properties; client geometry and renderer buffers are deliberate floating boundaries, not a promise of exact wide-integer pixel positions. There is no sparse/complex plotting implementation or numeric broadcasting promise here. Logical, scalar, vector, matrix, rank and empty admission follows each documented owner below rather than a universal coercion.
RuntimeOwned preparation retains the original BuiltinArguments through suspension and completion. Eligible direct GPU routes require the live owning provider, admitted precision/storage/shape, compatible exports and installed plotting device. NoAttempt means pre-effect inapplicability, not permission to swallow export, allocation, cancellation or emission errors. Existing provider errors are propagated and ingress errors are terminal Never replay. Source-property queries and figure persistence may require explicit owner-aware readback; surface property snapshots and quiver3/patch construction are not zero-copy routes.
Canonical UserInterface capability and host-callback effects reflect scene creation/presentation; function-sampling entries additionally declare Unknown effects for arbitrary user callbacks and require the Process execution stack. NativeAndWasm means the Runtime entry is portable, not that an absent graphics host is emulated. Portable Runtime linkage introduces no separate native numerical library artifact. Rendering is a fusion boundary and distributed execution is unsupported; no numeric provider kernel or fusion primitive is invented.
Chart scratch, metadata, retained render state and the sphere coordinate builder do not yet have complete finite host allocation ownership. The shared canonical HostAllocationReceipt policy guard refuses Managed work before extraction/gather/callbacks or scene mutation, preserves ordinary Unbounded routes, and checks captured cancellation generation and policy before publication. This is an explicit L25 gap, not allocation completion. Long synchronous geometry/copy loops, renderer/provider-private allocation accounting, full MATLAB property parity, physical performance and actual example execution remain independently required.
Supported representations and remaining forms
One or more X,Y,Z,C groups lower to independently addressable shared Patch objects, with leading axes and trailing shared properties retained. Vectors form one polygon; matched matrices split by columns, reusing row-count-compatible vectors. The result is a scalar handle or row of handles; standard 3-D view is applied even for z=0 polygons.
Shared Patch parsing owns FaceColor/EdgeColor, FaceAlpha/EdgeAlpha, LineWidth, DisplayName, Visible and color/indexed-CData behavior. Native X/Y/Z/C source class and shape remain authoritative. GPU coordinate/CData gathering precedes host polygon construction; no direct device polygon packer or full renderer color-mapping fidelity is fabricated. Typed-integer axes aliases retain their existing separate compatibility gate.
Examples
Create a red square in the z=0 plane
x = [0 1 1 0];
y = [0 0 1 1];
z = [0 0 0 0];
h = fill3(x, y, z, 'r');
get(h, 'Type')Expected output:
ans =
'patch'Create two filled polygons in one call
x1 = [0 1 0];
y1 = [0 0 1];
z1 = [0 0 0];
x2 = [2 3 3 2];
y2 = [0 0 1 1];
z2 = [1 1 1 1];
h = fill3(x1, y1, z1, 'g', x2, y2, z2, [0 0 1]);Apply patch properties to a fill3 polygon
x = [0 1 0];
y = [0 0 1];
z = [0.25 0.5 0.75];
h = fill3(x, y, z, [0.25 0.5 0.75], 'EdgeColor', 'none', 'FaceAlpha', 0.5);Create one patch per matrix column
X = [0 2; 1 3; 1 3; 0 2];
Y = [0 0; 0 0; 1 1; 1 1];
Z = [0 1; 0 1; 0 1; 0 1];
fill3(X, Y, Z, 'cyan');
view(3);Rendered fill3 example
x1 = [0 1.3 0.2];
y1 = [0 0.2 1.1];
z1 = [0.2 0.5 1.2];
x2 = [1.5 2.8 2.6 1.4];
y2 = [0.1 0.1 1.1 1.2];
z2 = [1.0 1.0 1.7 1.7];
x3 = [0.4 1.2 0.8];
y3 = [1.5 1.7 2.6];
z3 = [0.4 1.3 0.8];
fill3(x1, y1, z1, [0.2 0.7 0.95], x2, y2, z2, [0.95 0.45 0.2], x3, y3, z3, [0.4 0.85 0.45]);
title('3-D Polygon Assembly');
xlabel('x');
ylabel('y');
zlabel('z');
view(38, 25);
grid on;
Compare this result with other chart types in the plot examples.
Using fill3 with coding agents
Open a RunMat example with live inputs, then ask the agent to explain how fill3 changes the result.
Run a small fill3 example, explain the result, then change one input and compare the output.
FAQ
What does fill3 return?⌄
fill3 returns patch graphics handles. A single polygon group returns one scalar handle; multiple groups return a row vector of handles.
How is fill3 related to patch?⌄
fill3 lowers its positional X,Y,Z,C inputs into the same patch object machinery. This keeps geometry validation, color parsing, property handling, rendering, and handle behavior consistent with patch.
Can I pass patch properties to fill3?⌄
Yes. Trailing name-value pairs such as 'EdgeColor', 'none', 'FaceAlpha', 0.5, and 'LineWidth', 2 are forwarded to the patch objects created by fill3.
Does fill3 support GPU-resident inputs?⌄
GPU-resident coordinate inputs are gathered through the shared patch path for validation and polygon construction, then the resulting geometry is rendered through RunMat's GPU plotting renderer.
Does fill3 support indexed CData?⌄
Yes. Numeric scalar, vector, and matrix C inputs are retained as Patch CData. Renderer colormap fidelity continues to share the Patch renderer implementation.
Related Plotting functions
2D Charts
area · bar · errorbar · heatmap · hist · histogram · loglog · pie · plot · scatter · semilogx · semilogy · stairs · stem
Other
addpoints · ancestor · animatedline · autumn · axes · barh · bone · caxis · cla · clf · clim · colorcube · colororder · cool · copper · copyobj · cylinder · daspect · datacursormode · dataTipTextRow · drawnow · fcontour · figure · fill · findobj · fsurf · gobjects · gray · groot · hgload · hgsave · hidden · histogram2 · hold · hot · inferno · isgraphics · ishandle · jet · line · lines · linkaxes · magma · openfig · opengl · pan · parula · patch · pink · plasma · plotmatrix · plotyy · polarhistogram · polarplot · polarscatter · print · quiver3 · ribbon · saveas · savefig · sphere · spring · stackedplot · subtitle · summer · suptitle · text · textscatter · textscatter3 · triplot · turbo · viridis · waitbar · winter · wordcloud · xlabel · xlim · xline · xscale · xtickangle · xtickformat · xticklabels · xticks · ylabel · ylim · yline · yscale · ytickangle · ytickformat · yticklabels · yticks · zlim · zoom
More plotting resources
Source you can read
Every RunMat function's source is public. Read exactly how fill3 is executed, line by line, in Rust.
- View the source for fill3 in Rust on GitHub
- Learn how the RunMat runtime works
- Found a bug? Open an issue with a minimal reproduction.
About RunMat
RunMat is a runtime that executes MATLAB-syntax code fast on any GPU. The runtime is free for individuals, including freelancers; students and faculty at an academic institution, for teaching or research there; contributors to an open-source project, for work on that project; organizations with 250 or fewer people and US$10M or less in annual revenue; and anyone, for non-production use. See runmat.com/license.
- RunMat automatically optimizes your math for GPU execution on Apple, Nvidia, and AMD hardware. No code changes needed. Simulations that took hours now take minutes.
- Start running code in seconds. RunMat runs in the browser, on the desktop, or from the CLI. No license server, no IT ticket.