When a function returns a list, how do I break it apart so that I can access the parts that I want. For example, the function principal returns a lot of lists such as loading. Here's what I see using str():
..- attr(*, "dimnames")=List of 2
.. .. : chr [1:29] "A" "B" "C" "D" ... .. .. : chr [1:5] "RC1" "RC2" "RC3" "RC4" ...
How do I access the second and third loadings ("RC2" and "RC3") for the variable "C"?
loading is not a list. It's dimension names are a list, it is in fact a matrix. See below an example similar to yours using some example data in the package:
library(psych)
pc <- principal(Harman74.cor$cov,4,rotate="varimax")
str(pc)
#> List of 28
#> $ values : num [1:24] 8.14 2.1 1.69 1.5 1.03 ...
#> $ rotation : chr "varimax"
#> $ n.obs : logi NA
#> $ communality : Named num [1:24] 0.604 0.367 0.472 0.451 0.701 ...
#> ..- attr(*, "names")= chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> $ loadings : 'loadings' num [1:24, 1:4] 0.1566 0.0869 0.1428 0.2476 0.7861 ...
#> ..- attr(*, "dimnames")=List of 2
#> .. ..$ : chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> .. ..$ : chr [1:4] "RC1" "RC3" "RC2" "RC4"
#> $ fit : num 0.917
#> $ fit.off : num 0.968
#> $ fn : chr "principal"
#> $ Call : language principal(r = Harman74.cor$cov, nfactors = 4, rotate = "varimax")
#> $ uniquenesses: Named num [1:24] 0.396 0.633 0.528 0.549 0.299 ...
#> ..- attr(*, "names")= chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> $ complexity : Named num [1:24] 1.39 1.09 1.16 1.36 1.28 ...
#> ..- attr(*, "names")= chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> $ chi : num NA
#> $ EPVAL : num NA
#> $ R2 : Named num [1:4] 1 1 1 1
#> ..- attr(*, "names")= chr [1:4] "RC1" "RC3" "RC2" "RC4"
#> $ objective : num 2.17
#> $ residual : num [1:24, 1:24] 0.3961 -0.1408 -0.0972 -0.0346 0.024 ...
#> ..- attr(*, "dimnames")=List of 2
#> .. ..$ : chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> .. ..$ : chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> $ rms : num 0.0579
#> $ factors : int 4
#> $ dof : num 186
#> $ null.dof : num 276
#> $ null.model : num 11.4
#> $ criteria : Named num [1:3] 2.17 NA NA
#> ..- attr(*, "names")= chr [1:3] "objective" "" ""
#> $ PVAL : logi NA
#> $ weights : num [1:24, 1:4] -0.08029 -0.06611 -0.05358 -0.00913 0.24714 ...
#> ..- attr(*, "dimnames")=List of 2
#> .. ..$ : chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> .. ..$ : chr [1:4] "RC1" "RC3" "RC2" "RC4"
#> $ r.scores : num [1:4, 1:4] 1.00 -4.86e-17 -1.11e-15 -1.26e-15 -9.71e-17 ...
#> ..- attr(*, "dimnames")=List of 2
#> .. ..$ : chr [1:4] "RC1" "RC3" "RC2" "RC4"
#> .. ..$ : chr [1:4] "RC1" "RC3" "RC2" "RC4"
#> $ rot.mat : num [1:4, 1:4] 0.5991 -0.6178 -0.5092 -0.0105 0.4582 ...
#> $ Vaccounted : num [1:5, 1:4] 4.159 0.173 0.173 0.31 0.31 ...
#> ..- attr(*, "dimnames")=List of 2
#> .. ..$ : chr [1:5] "SS loadings" "Proportion Var" "Cumulative Var" "Proportion Explained" ...
#> .. ..$ : chr [1:4] "RC1" "RC3" "RC2" "RC4"
#> $ Structure : 'loadings' num [1:24, 1:4] 0.1566 0.0869 0.1428 0.2476 0.7861 ...
#> ..- attr(*, "dimnames")=List of 2
#> .. ..$ : chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> .. ..$ : chr [1:4] "RC1" "RC3" "RC2" "RC4"
#> - attr(*, "class")= chr [1:2] "psych" "principal"
loadings1<- pc$loadings # extract using name
# loadings1 is not a list, it is a matrix
str(loadings1)
#> 'loadings' num [1:24, 1:4] 0.1566 0.0869 0.1428 0.2476 0.7861 ...
#> - attr(*, "dimnames")=List of 2
#> ..$ : chr [1:24] "VisualPerception" "Cubes" "PaperFormBoard" "Flags" ...
#> ..$ : chr [1:4] "RC1" "RC3" "RC2" "RC4"
# If I want the second and third loadings for variable Cubes
# Here is one way to extract it
loadings1["Cubes", c("RC2", "RC3")]
#> RC2 RC3
#> 0.08144115 0.59303264
Created on 2020-06-12 by the reprex package (v0.3.0)
Thanks. That's helping me with my project.
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.