Journal article
Visualizing data through curvilinear representations of matrices
Computational statistics & data analysis, Vol.128, pp.255-270
12/2018
DOI: 10.1016/j.csda.2018.07.010
Abstract
Most high dimensional data visualization techniques embed or project the data onto a low dimensional space which is then used for viewing. Results are thus limited by how much of the information in the data can be conveyed in two or three dimensions. Methods11An R package implementing the proposed methodology is provided in the supplementary material.are described for a lossless functional representation of any real matrix that can capture key features of the data, such as distances and correlations. This approach can be used to visualize both subjects and variables as curves, allowing one to see patterns of subjects, patterns of variables, and how the subject and variable patterns relate to one another. A theoretical justification is provided for this approach, and various facets of the method’s usefulness are illustrated on both synthetic and real data sets.
•This paper provides a novel visualization method for any arbitrary matrix.•The proposed approach can be used to inspect the mean and the covariance structure.•There is great flexibility for atypical data formats such as dissimilarity matrices.•An R package implementing this visualization methodology is provided.
Details
- Title: Subtitle
- Visualizing data through curvilinear representations of matrices
- Creators
- Daniel K Sewell - Department of Biostatistics, University of Iowa, 145 N. Riverside Dr., Iowa City, IA 52242, 1-319-384-1585, United States
- Resource Type
- Journal article
- Publication Details
- Computational statistics & data analysis, Vol.128, pp.255-270
- Publisher
- Elsevier B.V
- DOI
- 10.1016/j.csda.2018.07.010
- ISSN
- 0167-9473
- eISSN
- 1872-7352
- Language
- English
- Date published
- 12/2018
- Academic Unit
- Biostatistics; Public Policy Center (Archive)
- Record Identifier
- 9984214790102771
Metrics
12 Record Views