Title:	Vizualizace vícerozměrných dat symbolovými grafy
Other Titles:	Visualization of multidimensional data using symbol plots
Authors:	Myslivec, Jaroslav Skalská, Hana
Citation:	E+M. Ekonomie a Management = Economics and Management. 2010, č. 3, s. 114-129.
Issue Date:	2010
Publisher:	Technická univerzita v Liberci
Document type:	článek article
URI:	http://www.ekonomie-management.cz/download/1331826750_9cdc/10_myslivec.pdf http://hdl.handle.net/11025/17359
ISSN:	1212-3609 (Print) 2336-5604 (Online)
Keywords:	vizualizace;symbolové grafy;atributy mapování;umístění symbolu
Keywords in different language:	visualization;symbol graphs;attributes mapping;symbol placement
Abstract in different language:	Visualization of multidimensional data is presented here as a tool for displaying quantitative content of data. The aim of visualization methods is to arrange information about the data file in such a way that new visual information is descriptive and informative enough and able to commu- nicate ideas easily and intuitively. With the using of graphical information it is possible to reach more profound understanding of the data and derive relevant information and knowledge from the data. This article is concerned with the categorization of visualization methods and brings extensive references related to this field. It focuses on selected symbol plots (profiles, polygons, face diagrams) in more details. The article describes fundamentals of these selected visualization techniques and also their advantages and disadvantages. It studies the characteristics of symbol plots when different mapping methods of attributes or different methods of placement of the sym- bols are used. Package “symbols” was designed and developed in the range of R language, which can change placement of symbols in the graph. This way it allows such an arrangement of symbols that is useful for visual identification of clusters with similar subjects. Data from public database is used here for the presentation of methods and possibilities of the software as well. This paper pre- sents new implementation which allows displaying symbols either in coordinate system of suitably chosen attributes or in coordinate system of two principal components. The new arrangement in placement of objects helps a user to discover associations and/or mutual relations in attributes, as well as to find clusters or structures of similar objects in data file.
Rights:	© Technická univerzita v Liberci CC BY-NC 4.0
Appears in Collections:	Číslo 3 (2010) Číslo 3 (2010)

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.