Lispix - Graphic Algorithm Tester with Tensors (Desmos like but Lisp) | The Rohans

Fri, 28 Aug 2026 00:00:00 +0000

▼ Output Quick Reference 7c108641

Core language
Form	Does
`(eval x)`	evaluate x (e.g. after quote)
`(quote x)` / `'x`	return x unevaluated, as-is
`(cons x y)`	construct pair (x . y)
`(car p)`	first element of pair/list p
`(cdr p)`	rest of list p
`(truncate n)`	integer part of n
`(< n1 n2)`	#t if n1 < n2, else ()
`(> n1 n2)`	#t if n1 > n2, else ()
`(consp x)`	#t if x is a non-empty list, cons, or closure
`(or x1 x2 ...)`	first x that is not (), else ()
`(and x1 x2 ...)`	last x if all are not (), else ()
`(not x)`	#t if x is (), else ()
`(cond (x1 y1) (x2 y2) ...)`	the first yi whose xi is non-()
`(if x y z)`	y if x is non-(), else z
`(let* (v1 x1) (v2 x2) ... y)`	sequentially bind v1..vk, then evaluate y
`(lambda v x)`	construct an anonymous closure
`(define v x)`	bind v to x, globally
`(defparameter v x)` / `(defvar v x)`	bind v to x, globally (CL style)
`(defun name (args) body)`	define a named function, globally
`(setq v x)`	update an existing binding in place
`(progn x1 x2 ... xk)`	evaluate in order, return the last
`(print x1 x2 ...)`	print each value to the Output pane, return the last

Math & tensors
`(+ ..)` `(- ..)` `(* ..)` `(/ ..)`	arithmetic; elementwise over tensors
`(sqrt x)` `(exp x)` `(log x)` `(tanh x)` `(sin x)` `(cos x)` `(abs x)`	elementwise math; scalar or tensor
`(pow t p)`	elementwise power
`(normalize v)`	unit vector
`(zero n)`	rank-1 tensor of n zeros
`(dot v1 v2)`	dot product
`(norm v)` / `(norm2 v)`	length / length²
`(dist v1 v2)` / `(dist2 v1 v2)`	distance / distance²
`(length lst)`	number of elements in a plain Lisp list
`(shape t)`	tensor of dimension sizes
`(rank t)`	number of dimensions (0 for scalars)
`(tensorp x)`	#t if x is a tensor
`(slice t i)`	element or row at index i
`(first t)` / `(rest t)`	first row / all rows after the first
`(slice-range t start end)`	rows/elements in [start, end)
`(col-slice M j)`	column j of a matrix, as a vector
`(reshape t [d0 d1 ...])`	reshape, same element count
`(matmul A B)` / `(@ A B)`	matrix product
`(transpose M)` / `(T M)`	swap rows and columns
`(vstack A B)`	stack A on top of B, row-wise
`(sum t)`	sum of all elements
`(amax t)`	maximum element
`(argmax t)`	index of the maximum element
`(softmax t)`	softmax; row-wise if rank 2
`(layer-norm t eps)`	layer normalization; row-wise if rank 2
`(causal-mask n)`	n×n lower-triangular attention mask
`(equalp v1 v2)`	#t if same shape and elementwise equal
`(make-tensor e1 e2 ...)`	backend for `[ ]` literals; rarely called directly

Drawing
`(point x y)`	draw a single point
`(line x0 y0 x1 y1)`	draw a line
`(circle x0 y0 r)`	draw a circle
`(color r g b)`	set the drawing color
`(background r g b)`	set the clear color
`(clear)`	fill the canvas with the background color
`(axes)` / `(axes tick-spacing)`	draw X/Y axis lines, ticks, numeric labels, and X/Y letters
`(set-origin 'graph)`	(0,0) at canvas center, Y up, X/Y can go negative
`(set-origin 'image)`	(0,0) at top-left, Y down, X/Y stay non-negative

Animation (inside `draw`)
`dt`	real seconds since the previous frame
`⏱️`	`dt` accumulated since page load — total elapsed seconds

Lispix — A Lisp Test Bed for Computer Vision, Machine Learning & Graphics Algorithms | The Rohans Limited on The Rohans