Parallel coordinates: Difference between revisions

Revision as of 08:05, 22 February 2020

Revision as of 17:56, 27 February 2020

Line 24:

Scaling is necessary because the plot is based on interpolation (linear combination) of consecutive pairs of variables.<ref name="Gpc2">{{cite book |first1=Rida |last1=Moustafa |first2=Edward J. |last2=Wegman |chapter=Multivariate continuous data – Parallel Coordinates |editors= Unwin, A.; Theus, M.; and Hofmann, H. (Eds.) |title=Graphics of Large Datasets: Visualizing a Million |publisher=Springer |pages=143–156 |year=2006 |isbn=978-0387329062 }}</ref> Therefore, the variables must be in common scale, and there are many scaling methods to be considered as part of data preparation process that can reveal more informative views.

A smooth parallel coordinate plot is achieved with splines.<ref name="Gpc1">{{cite journal |first1=Rida |last1=Moustafa |first2=Edward J. |last2=Wegman |title=On Some Generalizations of Parallel Coordinate Plots |journal=Seeing a Million, A Data Visualization Workshop, Rain Am Lech (nrNr.), Germany |year=2002 |url=http://herakles.zcu.cz/seminars/docs/infovis/papers/Moustafa_generalized_parallel_coordinates.pdf |archive-url=https://web.archive.org/web/20131224111246/http://herakles.zcu.cz/seminars/docs/infovis/papers/Moustafa_generalized_parallel_coordinates.pdf |url-status=dead |archive-date=2013-12-24 }}</ref> In the smooth plot, every observation is mapped into a parametric line (or curve), which is smooth, continuous on the axes, and orthogonal to each parallel axis. This design emphasizes the quantization level for each data attribute.<ref name="Gpc2" />

== Reading ==

Inselberg ({{harvnb|Inselberg|1997|p= }}) made a full review of how to visually read out parallel coords' relational patterns.<ref>{{citation|last1=Inselberg |first1=A.|year=1997 |chapter=Multidimensional detective |editor=|title=Information Visualization, 1997. Proceedings., IEEE Symposium on |series=|isbn=|place=|pages=100–107|chapter-url=httphttps://ieeexplore.ieee.org/xplsdocument/abs_all.jsp?arnumber=636793}}</ref> When most lines between two parallel axis are somewhat parallel to each other, it suggests a positive relationship between these two dimensions. When lines cross in a kind of superposition of X-shapes, it's a negative relationship. When lines cross randomly or are parallel, it shows there is no particular relationship.

== Limitations ==

Line 39:

== Software ==

While there are a large number of papers about parallel coordinates, there are only few notable software publicly available to convert databases into parallel coordinates graphics.<ref>{{cite web|url=http://eagereyes.org/techniques/parallel-coordinates|title=Parallel Coordinates|last=Kosara|first=Robert|year=2010}}</ref> Notable software are [[ELKI]], [[GGobi]], [[Macrofocus High-D]], [[Mondrian data analysis|Mondrian]], [[Orange (software)|Orange]] and [[ROOT]]. Libraries include [[Protovis.js]],<ref>{{cite web|url=https://mbostock.github.com/protovis/ex/cars.html|title=Protovis.js: Parallel Coordinates|last=Bostock|first=Mike|year=2011}}</ref> [[D3.js]]<ref>{{cite web|url=https://mbostock.github.com/d3/talk/20111116/iris-parallel.html|title=D3.js: Parallel Coordinates|last=Bostock|first=Mike|year=2012}}</ref><ref>{{cite web|url=http://bl.ocks.org/1341281|title=Parallel%20Coordinates|last=Davies|first=Jason|year=2011}}</ref> provide basic examples, while more complex examples are also available.<ref>{{cite web|url=http://exposedata.com/parallel/|title=Nutrient Contents - Parallel Coordinates|last=Chang|first=Kai|date=|year=2012|website=|url-status=dead|archive-url=https://web.archive.org/web/20160502023325/http://exposedata.com/parallel/|archive-date=2016-05-02|access-date=}}</ref><ref>http://bl.ocks.org/syntagmatic</ref><ref>{{Cite web|url=https://bl.ocks.org/IlievskiV/510869afe89b36eb46744cfbc2f1c1f1|title=Interactive exploring of the Laptop Prices dataset with Parallel Coordinates|last=Ilievski|first=Vladimir|date=2020-02-08|website=bl.ocks.org|url-status=live|archive-url=|archive-date=|access-date=2020-02-17}}</ref> D3.Parcoords.js<ref>{{cite web|url=https://syntagmatic.github.com/parallel-coordinates/|title=Parallel Coordinates (beta)|year=2012|last=Chang|first=Kai}}</ref> (a D3-based library) and [[Macrofocus High-D|Macrofocus High-D API]] (a Java library) specifically dedicated to parallel coordinates graphic creation have also been published. The [[Python (programming language)|Python]] data structure and analysis library [[Pandas (software)|Pandas]] implements parallel coordinates plotting, using the plotting library [[matplotlib]].<ref>[https://pandas.pydata.org/pandas-docs/version/0.21.0/visualization.html#parallel-coordinates Parallel Coordinates in Pandas]</ref> The [[R (programming language)|R]] programming language package [https://cran.r-project.org/web/packages/GGally/index.html GGally], among others, also implements parallel coordinates plotting.<ref>[https://cran.r-project.org/web/packages/GGally/index.html|title= Parallel coordinates in R]. </ref> High performance interactive parallel coordinates plots rendered with webgl can be made with the [[Plotly]] libraries in [https://plot.ly/python/parallel-coordinates-plot/ Python], [https://plot.ly/r/parallel-coordinates-plot/ R], and [https://plot.ly/javascript/parallel-coordinates-plot/ JavaScript].

== Other visualizations for multivariate data ==