Package to calculate the bidimensional regression between two 2D configurations following the approach by Tobler (1965).
The package is described in detail in: Carbon, C. C. (2013). BiDimRegression: Bidimensional Regression Modeling Using R. Journal of Statistical Software, Code Snippets, 52(1), 1-11, doi: 10.18637/jss.v052.c01.
library("devtools"); install_github("alexander-pastukhov/bidim-regression", dependencies=TRUE)
If you want vignettes, use
::install_github("alexander-pastukhov/tridim-regression", devtoolsdependencies=TRUE, build_vignettes = TRUE)
<- lm2(depV1 + depV2 ~ indepV1 + indepV2, NakayaData, 'euclidean') lm2euc <- lm2(depV1 + depV2 ~ indepV1 + indepV2, NakayaData, 'affine') lm2aff <- lm2(depV1 + depV2 ~ indepV1 + indepV2, NakayaData, 'projective') lm2prj anova(lm2euc, lm2aff, lm2prj predict(lm2euc) summary(lm2euc)
The author (CCC) is grateful to Waldo R. Tobler, now Professor Emeritus at the Department of Geography, University of California, Santa Barbara, for providing his original publications and his helpful correspondence and to Dan Montello for calling the author’s attention to Tobler’s work many years ago. I would like to thank Alinda Friedman, Gregory Francis, Jan de Leeuw, Achim Zeileis, two anonymous reviewers and Arne Terkowski for valuable comments on an earlier version of this paper, and Andrea Lyman and Vera M. Hesslinger for proofreading the manuscript. Last but not least, I am very indepted to Tomoki Nakaya, who has developed the original inference statistics of the overall models and the referring parameters and who helped me with reanalyzing these statistics as well as with ensuring the reliability of the used methods. Thank you!
All code is licensed under the GPL 3.0 license.