How this calculation works
Evaluate the components on a finite plane or spatial grid and inspect their raw values in the table. Arrows share one scale and colors encode relative sampled magnitude. The 3D view rotates through pointer dragging or labeled yaw/pitch sliders without changing the calculated vector data. Export the current SVG view, including its rotation.
F=(P,Q) or (P,Q,R); magnitude = √(P²+Q²+R²); use one shared arrow scale before orthographic projectionA worked example
For the planar field F(x,y)=(−y,x), the probe (1,1) has vector (−1,1) and magnitude √2. The arrows circulate around the origin, while the origin itself is a true zero vector. In 3D, adding R=z gives a spatial field whose projection changes as the view rotates.
Good to know
Finite samples of supported bounded arithmetic expressions; no integrated trajectories, automatic divergence/curl/potential proof or claim about behavior between grid points. Undefined samples are omitted and counted. Color uses the current sampled maximum, so different runs do not share an absolute color scale. Orthographic 3D projection can shorten nonzero vectors toward the viewing direction.
Reference: OpenStax Calculus: vector fields (opens in a new tab) ↗
A vector field assigns a direction and size to each position
In a planar field F(x,y)=(P,Q), the first component gives the horizontal part of the vector and the second gives its vertical part. In space, a third component R adds a direction along z. Each position supplies its own component values. The arrows are samples of that assignment, not objects that have already moved along a path.
Enter the component formulas separately with explicit multiplication. A 2D formula can use x and y; a 3D formula can also use z. Trigonometric functions take radians. A finite grid is evaluated locally in the browser and the table retains the raw coordinates, components and magnitude before any display scaling or projection.
Follow the rotational example through its actual components
For F(x,y)=(−y,x), the vector at (1,0) is (0,1), pointing upward. At (0,1) it is (−1,0), pointing left. At (−1,0) it is (0,−1), pointing downward. Those four positions illustrate a circulating pattern around the origin without claiming that the grapher has numerically integrated a circular trajectory.
At the probe (1,1), the vector is (−1,1) and its magnitude is √2. At the origin both components are zero, so a filled dot is used instead of an arrow with an invented direction. Increasing the distance from the origin increases the magnitude in this example. The arrows and colors encode that sampled variation using one common scale.
Use length and color without silently normalizing each arrow
Normalizing every vector to unit length would preserve direction but erase relative magnitude. This plot instead multiplies every component by one shared positive factor derived from the grid spacing and the largest sampled magnitude. A vector twice as large as another receives twice its pre-projection arrow length within that run.
The color scale also uses raw magnitude relative to the same sampled maximum, moving from blue toward red as the value increases. It is a relative scale for the current field and grid. Comparing a red arrow in one run with a red arrow in a different run does not establish that their original magnitudes are equal. Use the numerical table for that comparison.
Rotate a 3D field while keeping the projection limits visible
A spatial arrow has three components, but the screen has two coordinates. The viewer rotates the field with yaw and pitch and applies an orthographic projection. Drag inside the diagram to change that view, or use the labeled sliders for keyboard access. The underlying samples and their three-dimensional magnitudes remain unchanged.
An arrow pointing partly toward or away from the viewer looks shorter after projection. A nonzero vector almost aligned with the viewing direction can appear as an outlined dot; that is different from the filled dot used for a true zero vector. Magnitude colors still describe the raw vector rather than its shortened screen projection.
The current rotation is recorded in the SVG export. The cube edges and colored coordinate directions help orient the view, while the table remains the source of the actual x,y,z and P,Q,R numbers. A single projected view can hide spatial structure through overlap, so inspect several rotations before interpreting a dense diagram.
Keep undefined points and finite-grid limits explicit
A formula such as 1/x is undefined on x=0. If a grid includes that location, those samples are omitted and counted rather than replaced by zero arrows. The same rule applies when any component is outside its real domain or supported numerical range. A probe at an undefined point is reported as undefined without inventing a finite vector.
A grid can miss a narrow feature, a singularity between points or rapid variation. Increasing density changes the sampled picture but does not prove continuity or establish a global field property. The 3D limit of seven samples per axis keeps the view bounded at 343 vectors; the plane supports up to fifteen per axis. More arrows are not automatically a clearer explanation.
Separate a field picture from derivative and trajectory calculations
A source-like radial pattern, a sink-like inward pattern and a circulating pattern are useful visual starting points. Divergence, curl and conservativeness are mathematical properties requiring appropriate derivatives and domain assumptions. The plot does not infer them merely from the direction of a few arrows. A conservative field is associated with a potential, and checking that statement requires more than a plausible-looking image.
The polynomial vector field divergence and curl calculator performs a separate analytic derivative task for supported planar polynomials. The first-order ODE solver instead constructs numerical initial-value trajectories for a scalar rate equation and shows its slope field. Those tools answer different questions; neither a field arrow nor its visual length alone supplies a full solution curve.
Keep the component formulas, grid range, density, probe and view settings with a shared diagram. The copied scenario and SVG preserve those choices, while the table makes clear that the drawing shows sampled vectors rather than measured physical velocities or an automatically fitted wind, gravity or fluid model.
A few questions, answered
Can I enter nonpolynomial components?
Yes. The bounded arithmetic grammar includes sin, cos, tan, exp, natural log, sqrt and abs, with explicit multiplication and permitted coordinate variables.
Does rotating the view change the field?
No. It changes only the orthographic projection; component values and raw magnitudes remain the same.
Why can a nonzero 3D vector look very short?
It can point nearly toward the viewing direction. Projection shortens its screen arrow while its raw magnitude remains in the color and table.
Are undefined grid points drawn as zero vectors?
No. They are omitted and counted separately.
Does this plot prove that a field is conservative?
No. That requires mathematical analysis, including the field’s derivatives and domain.