Interesting examples to reproduce

[tlienart]

Opinionated list of examples that

• https://jsxgraph.org/wiki/index.php/Draggable_exponential_function
  • doable
• https://jsxgraph.org/wiki/index.php/Lagrange_interpolation
  • requires button
• https://jsxgraph.org/wiki/index.php/Least-squares_circle_fitting
  • requires circle
• https://jsxgraph.org/wiki/index.php/Least-squares_line_fitting
  • requires line and circle
• https://jsxgraph.org/wiki/index.php/Lituus
  • doable
• https://jsxgraph.org/wiki/index.php/Logarithmic_spiral
  • tangent, glider
• https://jsxgraph.org/wiki/index.php/Matrix_multiplication , https://jsxgraph.org/wiki/index.php/Matrix_multiplication_II
  • arrows
• https://jsxgraph.org/wiki/index.php/P-Norm
  • doable
• https://jsxgraph.org/wiki/index.php/Polynomial_regression_I
  • glider, text
• https://jsxgraph.org/wiki/index.php/Polynomial_regression_II
  • text
• https://jsxgraph.org/wiki/index.php/Polynomials
  • doable
• https://jsxgraph.org/wiki/index.php/Power_Series_for_sine, https://jsxgraph.org/wiki/index.php/Power_Series_for_sine_and_cosine https://jsxgraph.org/wiki/index.php/Power_series_for_the_exponential_function
  • doable
• https://jsxgraph.org/wiki/index.php/Riemann_integrability , https://jsxgraph.org/wiki/index.php/Riemann_sum_II, https://jsxgraph.org/wiki/index.php/Riemann_sum_III https://jsxgraph.org/wiki/index.php/Riemann_sums
  • riemansum
• https://jsxgraph.org/wiki/index.php/Inequalities
  • inequality, line
• https://jsxgraph.org/wiki/index.php/Hilbert_curve
  • doable but they use a horrible function to do it...

(haven't checked what's before letter D)

• https://jsxgraph.org/wiki/index.php/Differential_equations

Other stuffs / docs

• https://jsxgraph.org/docs/symbols/Button.html

[tlienart]

https://tlienart.github.io/JSXGraphWeb/page2/

[s-celles]

Note: This recap was generated by an AI agent (Claude Code) acting on my behalf — please flag any inaccuracies. Each "Done" item below has been independently verified against the 0.5.3 source (commit 338f5b9), not trusted from the 2020 [x] marks.

Status as of v0.5.3

✅ Ported to docs/src/wiki_examples.md

Seven examples from this checklist now live on the new Wiki Examples documentation page (deployed once CI finishes). Each one is a self-contained @example block that renders an interactive board.

#	Wiki page	Section in the docs	Notes
1	Draggable exponential function	§1	Uses the wiki's A.Y()^(x/A.X()) formula so the anchor point is on the curve by construction. Requires v0.5.3's @jsf property-access (α.X(), α.Y()) support.
2	Lituus	§2	Slider k, polar curve √(k/φ). Built as a parametric curve(fx, fy, t_start, t_end) because the standalone polar returns its own board.
3	P-Norm	§3	Slider p, implicit curve `
4	Polynomials	§4	Four-slider quartic a·x⁴ + b·x³ + c·x² + d·x. Wiki comparison pending.
5	Power Series for sine	§5	Slider N for truncation order. Body is a hand-written JSXGraph.JSFunction because @jsf does not transpile for/while (:for is in UNSUPPORTED_EXPR_HEADS); the placeholder marker __JSF_REF_s__ is substituted at render time via the same mechanism @jsf uses internally.
6	Riemann sums	§6	Three sliders (n, start, end). Wraps each slider in @jsf () -> slider_var so JSXGraph receives function(){return el_xxx.Value();} — required because functiongraph and riemannsum range slots don't accept raw slider element refs.
7	Inequalities	§7	Linear half-planes. Adds a new public method line(c::Real, b::Real, a::Real) for the JSXGraph homogeneous-coords form c + b·x + a·y = 0.

🍒 Lot 2 — low-hanging fruits (~30 min total)

Direct variants of work already done; no new constructors, no engine changes.

Wiki page	Pattern to reuse	Tweak
Power Series for sine and cosine	§5 — raw JSFunction with truncation slider	Add a second JSFunction with cos terms (-1)^k · t^(2k) / (2k)!
Power series for the exponential	§5	Even simpler: t^k / k!, no sign alternation
Riemann sum II / III, Riemann integrability	§6 — three-slider sum	Swap f, the interval defaults, and the type ("right", "trapezoidal", …)

🍐 Lot 3 — moderate (~15–20 min each)

Wider in scope but still reachable with the v0.5.3 API.

Wiki page	What it needs
Matrix multiplication I	4 sliders for matrix entries + 1 draggable input vector + arrow for the output vector. Output uses @jsf (x, y) -> a*x + b*y style functions reading vec.X() / vec.Y().
Hilbert curve	Recurse in Julia to produce Vector{Float64} of (x, y) pairs, then feed a data-driven curve through the Tables.jl extension: plot((x=…, y=…), :x, :y). Static (no slider).
Logarithmic spiral	Pattern from §2 (Lituus parametric form) + a glider on the spiral + a tangent element through that glider.

🍋 Lot 4+ — defer until a Lot 3 follow-up

Each of these needs either a JS regression call (JXG.Math.Numerics.*), a mutable button-triggered state machine, or a full RK4 solver in raw JS. Worth tracking in this issue but not low-effort.

Wiki page	Blocker
Lagrange interpolation	button controlling a mutable point list; need to manage state in JS.
Polynomial regression I, II	Calls JXG.Math.Numerics.regression… from inside a raw JSFunction.
Least-squares circle, line fitting	Same blocker as Polynomial regression.
Matrix multiplication II	The extended variant adds eigenvector visualization — wider scope.
Differential equations	Hand-written RK4 in a JSFunction body.

🟥 Known limitations exposed during this work

• @jsf does not transpile for / while (see UNSUPPORTED_EXPR_HEADS in src/jsfunction.jl). Workaround: build a raw JSXGraph.JSFunction(code; refs=…) where the body uses placeholders like __JSF_REF_s__ for captured element refs (§5 is the reference example).
• functiongraph only exposes the single-parent form functiongraph(f). To pass xmin / xmax as functions or sliders, you currently have to call JSXGraph._create_element("functiongraph", (f, xmin, xmax), …) (used in §6). A public functiongraph(f, xmin, xmax) constructor would be worth adding.
• JSXGraph's inequality element only accepts a line parent — for a curve-based inequality (y ≤ f(x) for an arbitrary f) the upstream API doesn't exist, so the "y ≤ sin(x)" framing in earlier drafts of this page was a stretch.

Suggested next step

Pick up Lot 2 (the three sine/cos/exp series and a Riemann variant) as a single docs-only follow-up — purely additive, no API risk. Lot 3 is the right next bite once that lands.