Abstract

Automatic labeling in visualization
◼
  • Where is automatic labeling used?
    • ◼
  • Why automation?
    • ◼
  • What is automated?
    • ◼
  • How is it automated?
    • ◼
  • When was it rolled out?

    • Where is labeling used?

    Applications of automatic labeling

    ◼
  • Map
    • GeoGraphics
    [source]
    [source]
    ◼
  • Data Analysis
    • [source]
    ◼
  • Scientific Plots
    • [source]
    ◼
  • Anatomy
    

    • Why Automation?

    ◼
  • Free-hand labeling can be time consuming and hard to reproduce.
    
    • ◼
  • Without an algorithmic method, labels can collide with each other and become hard to read.
    
    • ◼
  • Using low level Text to add labels requires post-processing of the plots.
    
    • ◼
  • Free-hand labeling can distract from the process of data exploration.
    

    • What is automated?

    Data

    ◼
  • Categorical data
    • ◼
  • Meta data

    • Options

    ◼
  • PlotLabels
    • ◼
  • LabelingFunction
    • ◼
  • PlotHighlighting
    • ◼
  • ChartLabels

    • Wrapper

    ◼
  • Callout
    • ◼
  • Labeled
    • ◼
  • Highlighted

    • Categorical data

    ◼
  • DateListStepPlot
    
    • ◼
  • BubbleChart
    • In[]:=
    {{Apple type, size, price}, …}
    Out[]=
    {{{Red Delicious,72s,36.},{Red Delicious,88s,34.5},{Red Delicious,100s,34.5},{Red Delicious,113s,32.5},{Red Delicious,125s,34.5}},{{Golden Delicious,72s,46.5},{Golden Delicious,80s,45.},{Golden Delicious,88s,43.5},{Golden Delicious,100s,39.5},{Golden Delicious,113s,38.5},{Golden Delicious,125s,35.5}},{{Fuji,64s,48.5},{Fuji,72s,48.},{Fuji,88s,56.},{Fuji,100s,34.5},{Fuji,113s,44.}},{{Granny Smith,64s,46.5},{Granny Smith,72s,46.5},{Granny Smith,88s,54.},{Granny Smith,100s,38.5},{Granny Smith,113s,36.5}},{{Gala,72s,42.},{Gala,80s,42.},{Gala,88s,44.},{Gala,100s,39.5},{Gala,113s,39.},{Gala,125s,38.5}},{{Honeycrisp,56s,92.5},{Honeycrisp,64s,62.5},{Honeycrisp,72s,75.5},{Honeycrisp,80s,80.}},{{Pink Lady,64s,56.},{Pink Lady,72s,42.},{Pink Lady,80s,41.5},{Pink Lady,113s,42.}}}
    In[]:=
    BubbleChart
    {{Apple type, size, price}, …}
    ,ScalingFunctions->{NominalScale[Automatic],NominalScale[{"125s","113s","100s","88s","80s","72s","64s","56s"}],None},AspectRatio12,FrameLabel->{{"size",None},{"type",None}},ImageSize->600
    ◼
  • BarChart
    
    • ◼
  • ParallelAxisPlot
    • In[]:=
    orange data
    //Short
    Out[]//Short=
    {{ATLANTA,FLORIDA,4/5 bushel cartons,NAVEL,56s,U.S. No. 1},236,{SAN FRANCISCO,MEXICO,7/10 bushel cartons,VALENCIA,113s,No Grade Marks}}
    In[]:=
    ParallelAxisPlot
    orange data
    ,ScalingFunctionsNominalScale[Automatic],AxesLabel->{"Location","Origin","Package","Variety","Item Size","Grade"},ImageSize1000,LabelStyle14
    Out[]=

    Highlighting labels
    

    Line Labeling
    

    1D Point Labeling
    

    2D and 3D Point Labeling

    ListPlot[RandomReal[1,{50,2}]RandomWord[50]]
    Out[]=
    Table​​ListPlotRandomVariate[NormalDistribution[0,σ],{100,2}]->RandomWord[100],​​
    ​​,​​{σ,{0.2,0.5,1}}​​
    Out[]=
    
    ,
    ,
    
    In[]:=
    FeatureSpacePlotEntityValue
    Europe
    COUNTRIES
    ,"FlagImage",LabelingFunctionCallout
    Out[]=
    In[]:=
    Table​​FeatureSpacePlotEntityValue
    Europe
    COUNTRIES
    ,"FlagImage",​​LabelingFunctionCallout,​​LabelingSizesize,​​ImageSize360,​​BackgroundLightGray,​​{size,{20,60}}​​
    Out[]=
    
    ,
    
    Out[]=
    Out[]=
    
    ,
    ,
    

    Parametric Curve Labeling
    

    Region Labeling
    

    Surface Labeling
    

    Labeling In Charts

    Out[]=
    Out[]=
    label1
    label2
    label3
    Out[]=
    In[]:=
    TableBarChart{1,1,2,3,5,8,13},​​ChartLabels->
    ,
    ,
    ,
    ,
    ,
    ,
    ,​​ImageSizesize,​​LabelingSizeAutomatic,​​BaselinePosition->Axis​​,​​{size,{200,400}}​​
    Out[]=
    Out[]=

    How?

    Methods of implementation of automatic labeling

    Image Processing
    

    Optimization
    

    Geometry
    

    Point Feature Labeling Problem
    

    Others
    

    When is it rolled out?

    ◼
  • Timeline of labeling
    • In[]:=
    TimelinePlot[data,​​ImageSize1100,​​PlotRangePadding{{Scaled[0.1],Scaled[0.1]},None},​​AspectRatio1/1.75,​​PlotStyleColorData[108]​​]​​
    Out[]=
    data
    

    Showcases

    In[]:=
    l=
    all countries, dependencies, and territories
    COUNTRIES
    ["Position"];​​c=
    all countries, dependencies, and territories
    COUNTRIES
    ["Name"];​​oil=
    all countries, dependencies, and territories
    COUNTRIES
    
    petroleum
    ;​​major=Quantile[QuantityMagnitude@DeleteMissing[oil],0.75];​​oildata=Select[Select[Transpose[{l,oil,c}],Head@#[[2]]=!=Missing&],QuantityMagnitude[#[[2]]]>major&];​​GeoBubbleChart[oildata[[All,1;;2]],ColorFunction"Rainbow",LabelingFunction(Callout[oildata[[All,3]][[#2[[2]]]],LabelStyle13]&),ImageSize1200,PlotTheme->"Marketing",GeoProjection"Albers"]
    Out[]=

    Callout Appearance

    (*markup*)​​Callout[data,"label",​​LabelStyle40,​​LeaderSize{{leaderlength,angle,gap},{necklength,angle}},​​Appearanceappearance,​​CalloutMarkermarker,​​Backgroundbackgroundstyle,​​FrameMarginsframemargins,​​RoundingRadiusradius,​​CalloutStyleThickness[0.01]]
    init
    

    CITE THIS NOTEBOOK

    Wolfram R&D LIVE: Labeling Everywhere​
    by MinHsuan Peng​
    Wolfram Community, STAFF PICKS, September 6, 2023
    ​https://community.wolfram.com/groups/-/m/t/3007543