9.3
62 Probability and Statistical Diagrams
Small, immutable, addressable diagram groups support common probability and statistics explanations. These constructors do not analyse data or add a separate scene protocol: their bars, cells, tree branches, quartile box, and error-bar elements are ordinary named children that work with existing paths, colours, transforms, and attention effects.
procedure
(bar-chart values #:id identifier [ #:labels labels #:center center #:width width #:height height #:maximum maximum #:fill fill #:stroke stroke #:stroke-width stroke-width #:value-labels? value-labels?]) → group-visual? values : (and/c list? pair?) identifier : symbol? labels : (or/c (listof string?) false/c) = #f center : vec2? = origin width : positive-real? = 6 height : positive-real? = 3 maximum : (or/c positive-real? false/c) = #f fill : any/c = "cornflowerblue" stroke : any/c = "navy" stroke-width : nonnegative-real? = 2 value-labels? : boolean? = #t
Creates a baseline and one upward nonnegative bar per value. Values are scaled
against maximum, or the largest supplied value (at least one). Labels
and value labels are optional ordinary text children. The path of one bar is
(bar-chart-bar-path identifier index), where indexes start at one.
procedure
(histogram samples #:id identifier [ #:bins bins #:range range #:center center #:width width #:height height]) → group-visual? samples : (and/c list? pair?) identifier : symbol? bins : exact-positive-integer? = 8 range : (or/c pair? false/c) = #f center : vec2? = origin width : positive-real? = 6 height : positive-real? = 3
Counts finite numeric samples into equally wide bins, then returns the same
addressable bar-group shape as bar-chart. A supplied range is an
increasing pair; samples outside it are omitted and a sample at the upper bound
belongs to the final bin.
procedure
(stacked-bar-chart rows #:id identifier [ #:center center #:width width #:height height #:maximum maximum #:colors colors #:stroke stroke #:stroke-width stroke-width]) → group-visual? rows : (and/c list? pair?) identifier : symbol? center : vec2? = origin width : positive-real? = 6 height : positive-real? = 3 maximum : (or/c positive-real? false/c) = #f colors : (and/c list? pair?) = any/c stroke : any/c = "navy" stroke-width : nonnegative-real? = 1
Creates one bar per equal-length nonnegative row and stacks each row’s values
from the common baseline. Use stacked-bar-path and
stacked-bar-segment-path for one bar or segment, with one-based
indexes.
procedure
(sample-space rows #:id identifier [ #:center center #:width width #:height height]) → group-visual? rows : (and/c list? pair?) identifier : symbol? center : vec2? = origin width : positive-real? = 5 height : positive-real? = 3
Creates equally sized, coloured cells from a nonnegative rectangular matrix.
Each cell displays its supplied weight. Its path is
(sample-space-cell-path identifier row column), with one-based row and
column indexes; the geometry remains equal even for unequal weights.
struct
(struct probability-branch (id label probability children) #:transparent) id : symbol? label : string? probability : nonnegative-real? children : (listof probability-branch?)
Describes one immutable node in a finite probability tree. Branch identities
must be globally unique within one probability-tree input.
procedure
(probability-tree branches #:id identifier [ #:center center #:width width #:level-gap gap #:node-radius radius]) → group-visual? branches : (and/c list? pair?) identifier : symbol? center : vec2? = origin width : positive-real? = 6 gap : positive-real? = 1 radius : positive-real? = 1/6
Lays out an explicit finite forest by leaf order. Child edge labels display the
child branch probability. A node can be addressed with
(probability-tree-node-path identifier branch-id).
procedure
(box-plot values #:id identifier [ #:center center #:width width #:height height]) → group-visual? values : (listof finite-real?) identifier : symbol? center : vec2? = origin width : positive-real? = 5 height : positive-real? = 3/4
Creates whiskers, a quartile box, and a median line from at least two values.
Quartiles use deterministic linear interpolation between sorted observations;
outliers are not inferred or displayed. box-plot-summary is the
transparent five-number summary structure used by the constructor.
struct
(struct error-bar-point (x y error) #:transparent) x : finite-real? y : finite-real? error : nonnegative-real?
Describes a point with symmetric vertical error.
procedure
(error-bars points #:id identifier [ #:center center #:cap-width cap-width]) → group-visual? points : (and/c list? pair?) identifier : symbol? center : vec2? = origin cap-width : positive-real? = 1/5
Creates addressable vertical stems, end caps, and point markers. The path for a
one-based point index is (error-bar-path identifier index).
procedure
(bar-chart-bar-path identifier index) → visual-path?
identifier : symbol? index : exact-positive-integer?
Returns the path of one one-based bar.
procedure
(stacked-bar-path identifier index) → visual-path?
identifier : symbol? index : exact-positive-integer?
Returns the path of one one-based stacked bar.
procedure
(stacked-bar-segment-path identifier bar-index segment-index) → visual-path? identifier : symbol? bar-index : exact-positive-integer? segment-index : exact-positive-integer?
Returns the path of one one-based stacked-bar segment.
procedure
(sample-space-cell-path identifier row column) → visual-path? identifier : symbol? row : exact-positive-integer? column : exact-positive-integer?
Returns the path of one one-based sample-space cell.
procedure
(probability-tree-node-path identifier branch-id) → visual-path? identifier : symbol? branch-id : symbol?
Returns the path of one named probability-tree node.
struct
(struct box-plot-summary ( minimum lower-quartile median upper-quartile maximum) #:transparent) minimum : finite-real? lower-quartile : finite-real? median : finite-real? upper-quartile : finite-real? maximum : finite-real?
The deterministic five-number summary used by box-plot.
procedure
(error-bar-path identifier index) → visual-path?
identifier : symbol? index : exact-positive-integer?
Returns the path of one one-based error-bar group.
See "examples/probability-and-statistics.rkt" for one composition that uses chart, finite-outcome, tree, distribution, and uncertainty views together. }