#SOM example using wines data set
library(kohonen)
## Warning: package 'kohonen' was built under R version 3.5.3
search()
## [1] ".GlobalEnv" "package:kohonen" "package:stats"
## [4] "package:graphics" "package:grDevices" "package:utils"
## [7] "package:datasets" "package:methods" "Autoloads"
## [10] "package:base"
ls(2)
## [1] "add.cluster.boundaries" "check.whatmap"
## [3] "classmat2classvec" "classvec2classmat"
## [5] "expandMap" "getCodes"
## [7] "map" "object.distances"
## [9] "som" "somgrid"
## [11] "supersom" "tricolor"
## [13] "unit.distances" "xyf"
help("som")
## starting httpd help server ... done
help("cutree")
data(wines)
str(wines)
## num [1:177, 1:13] 13.2 13.2 14.4 13.2 14.2 ...
## - attr(*, "dimnames")=List of 2
## ..$ : NULL
## ..$ : chr [1:13] "alcohol" "malic acid" "ash" "ash alkalinity" ...
dim(wines)
## [1] 177 13
wineskala<-scale(wines)
grid=somgrid(5,5,"hexagonal")
grid
## $pts
## x y
## [1,] 1.5 0.8660254
## [2,] 2.5 0.8660254
## [3,] 3.5 0.8660254
## [4,] 4.5 0.8660254
## [5,] 5.5 0.8660254
## [6,] 1.0 1.7320508
## [7,] 2.0 1.7320508
## [8,] 3.0 1.7320508
## [9,] 4.0 1.7320508
## [10,] 5.0 1.7320508
## [11,] 1.5 2.5980762
## [12,] 2.5 2.5980762
## [13,] 3.5 2.5980762
## [14,] 4.5 2.5980762
## [15,] 5.5 2.5980762
## [16,] 1.0 3.4641016
## [17,] 2.0 3.4641016
## [18,] 3.0 3.4641016
## [19,] 4.0 3.4641016
## [20,] 5.0 3.4641016
## [21,] 1.5 4.3301270
## [22,] 2.5 4.3301270
## [23,] 3.5 4.3301270
## [24,] 4.5 4.3301270
## [25,] 5.5 4.3301270
##
## $xdim
## [1] 5
##
## $ydim
## [1] 5
##
## $topo
## [1] "hexagonal"
##
## $neighbourhood.fct
## [1] bubble
## Levels: bubble gaussian
##
## $toroidal
## [1] FALSE
##
## attr(,"class")
## [1] "somgrid"
som.wines<-som(scale(wines),grid=somgrid(5,5,"hexagonal"))
summary(som.wines)
## SOM of size 5x5 with a hexagonal topology and a bubble neighbourhood function.
## The number of data layers is 1.
## Distance measure(s) used: sumofsquares.
## Training data included: 177 objects.
## Mean distance to the closest unit in the map: 3.507.
str(som.wines)
## List of 13
## $ data :List of 1
## ..$ : num [1:177, 1:13] 0.255 0.206 1.702 0.305 1.491 ...
## .. ..- attr(*, "dimnames")=List of 2
## .. .. ..$ : NULL
## .. .. ..$ : chr [1:13] "alcohol" "malic acid" "ash" "ash alkalinity" ...
## .. ..- attr(*, "scaled:center")= Named num [1:13] 12.99 2.34 2.37 19.52 99.59 ...
## .. .. ..- attr(*, "names")= chr [1:13] "alcohol" "malic acid" "ash" "ash alkalinity" ...
## .. ..- attr(*, "scaled:scale")= Named num [1:13] 0.809 1.119 0.275 3.336 14.174 ...
## .. .. ..- attr(*, "names")= chr [1:13] "alcohol" "malic acid" "ash" "ash alkalinity" ...
## $ unit.classif : num [1:177] 15 24 25 23 25 15 23 20 20 24 ...
## $ distances : num [1:177] 3.578 2.901 3.027 1.586 0.697 ...
## $ grid :List of 6
## ..$ pts : num [1:25, 1:2] 1.5 2.5 3.5 4.5 5.5 1 2 3 4 5 ...
## .. ..- attr(*, "dimnames")=List of 2
## .. .. ..$ : NULL
## .. .. ..$ : chr [1:2] "x" "y"
## ..$ xdim : num 5
## ..$ ydim : num 5
## ..$ topo : chr "hexagonal"
## ..$ neighbourhood.fct: Factor w/ 2 levels "bubble","gaussian": 1
## ..$ toroidal : logi FALSE
## ..- attr(*, "class")= chr "somgrid"
## $ codes :List of 1
## ..$ : num [1:25, 1:13] -1.0016 0.0419 0.3961 0.1893 0.5157 ...
## .. ..- attr(*, "dimnames")=List of 2
## .. .. ..$ : chr [1:25] "V1" "V2" "V3" "V4" ...
## .. .. ..$ : chr [1:13] "alcohol" "malic acid" "ash" "ash alkalinity" ...
## $ changes : num [1:100, 1] 0.0522 0.056 0.0525 0.0546 0.0546 ...
## $ alpha : num [1:2] 0.05 0.01
## $ radius : Named num [1:2] 3 0
## ..- attr(*, "names")= chr [1:2] "66.66667%" ""
## $ user.weights : num 1
## $ distance.weights: num 1
## $ whatmap : int 1
## $ maxNA.fraction : int 0
## $ dist.fcts : chr "sumofsquares"
## - attr(*, "class")= chr "kohonen"
som.wines$unit.classif
## [1] 15 24 25 23 25 15 23 20 20 24 15 15 25 25 24 24 23 25 19 20 19 15 15
## [24] 19 23 15 15 23 15 24 24 15 23 23 19 23 15 15 19 20 14 20 14 20 14 20
## [47] 20 24 25 20 24 25 24 20 24 20 24 25 16 1 2 16 17 1 17 16 16 2 21
## [70] 9 18 7 22 17 16 16 8 21 13 16 17 6 3 17 16 6 6 6 6 1 1 1
## [93] 12 17 22 22 16 17 12 16 16 7 11 11 7 11 7 11 18 12 11 1 6 7 6
## [116] 11 7 8 11 12 18 13 13 12 11 7 6 7 8 9 9 9 9 2 4 3 3 3
## [139] 2 2 9 2 3 9 9 3 3 4 5 5 5 5 4 4 3 4 4 5 5 4 2
## [162] 2 9 4 3 5 4 5 5 9 4 4 3 5 4 4 5
plot(som.wines)
som.wines$grid$pts
## x y
## [1,] 1.5 0.8660254
## [2,] 2.5 0.8660254
## [3,] 3.5 0.8660254
## [4,] 4.5 0.8660254
## [5,] 5.5 0.8660254
## [6,] 1.0 1.7320508
## [7,] 2.0 1.7320508
## [8,] 3.0 1.7320508
## [9,] 4.0 1.7320508
## [10,] 5.0 1.7320508
## [11,] 1.5 2.5980762
## [12,] 2.5 2.5980762
## [13,] 3.5 2.5980762
## [14,] 4.5 2.5980762
## [15,] 5.5 2.5980762
## [16,] 1.0 3.4641016
## [17,] 2.0 3.4641016
## [18,] 3.0 3.4641016
## [19,] 4.0 3.4641016
## [20,] 5.0 3.4641016
## [21,] 1.5 4.3301270
## [22,] 2.5 4.3301270
## [23,] 3.5 4.3301270
## [24,] 4.5 4.3301270
## [25,] 5.5 4.3301270
som.wines$codes #data code book vector
## [[1]]
## alcohol malic acid ash ash alkalinity magnesium
## V1 -1.00157078 -0.55562243 0.41339214 0.02299006 -0.42341183
## V2 0.04190582 0.02550588 0.30541177 0.19661484 0.08986299
## V3 0.39611452 2.14437120 -0.01717473 0.54307656 -0.72519841
## V4 0.18931086 0.58084045 -0.04108536 0.37742052 -0.22066427
## V5 0.51566355 0.65373389 0.86072119 1.17976308 0.59099503
## V6 -1.55346525 -0.77783381 0.32148015 1.22742882 -1.12192425
## V7 -0.57858510 -0.45176858 -0.13358797 0.89933508 -0.62229509
## V8 -0.85503371 1.03192514 -0.49762172 -0.13997962 -0.50359346
## V9 -0.29046281 0.44759729 -0.19382654 0.15461269 0.18906035
## V10 0.32683085 0.32090320 0.10427186 0.13401453 0.20716561
## V11 -1.03841922 -0.31952949 -1.18531232 0.13973407 -0.96103721
## V12 -1.35934903 0.79109996 -0.57303607 -0.04659604 -0.50079450
## V13 -0.26659780 1.89281728 0.17345848 0.90980683 -0.19854491
## V14 0.90382935 1.19818651 -0.15188036 -0.34348605 0.22154147
## V15 0.73886262 -0.56394514 -0.08556666 -1.05614308 -0.22611302
## V16 -0.62109416 -0.96731259 -1.92239332 -0.86780742 -0.80373798
## V17 -0.99473109 -0.91740118 -0.15691498 -0.24310251 -0.44026539
## V18 -0.72690664 -0.54563009 1.59010400 1.26386381 0.02135997
## V19 0.75551765 0.47927149 0.59991127 -0.63569638 0.72957321
## V20 1.06274190 -0.49006841 -0.43938104 -1.15423202 0.65669531
## V21 -0.84954952 -1.02347755 -1.59413105 -0.88527362 2.01731698
## V22 -0.64786727 -0.54193135 0.53837753 0.99357712 2.72923140
## V23 0.67326766 -0.34080017 1.58252166 0.12215951 1.16758347
## V24 0.91320647 -0.42193169 0.79844068 -0.24974151 0.52421916
## V25 1.45971609 -0.55807548 0.22790343 -1.31120908 0.46956994
## tot. phenols flavonoids non-flav. phenols proanth col. int.
## V1 -0.9400140 -0.43936415 1.403295008 -0.61564872 -0.81273220
## V2 -0.5448729 -1.23023067 1.448069047 -1.42035007 -0.02019294
## V3 -1.1597088 -1.37015099 1.368631305 -0.89189478 0.36125989
## V4 -0.9581536 -1.33091980 1.151808110 -0.61838348 1.69078016
## V5 -0.7493124 -0.88968489 0.058251838 -0.15059388 1.90595031
## V6 -0.2448057 -0.19888750 0.744305578 -0.09264421 -1.12432326
## V7 -0.2738831 0.11736391 0.342886464 -0.25017775 -0.91904906
## V8 -0.6975406 -0.55697910 0.451368469 -0.83818288 -0.82792585
## V9 -1.4965649 -1.18047424 -0.304326552 -1.14154603 0.08095311
## V10 -0.5762809 -0.63422641 -0.037715820 -0.49014654 0.69603171
## V11 -0.1126174 -0.02739517 -0.146084327 -0.03309808 -1.12842151
## V12 0.8563408 0.67598083 -0.552537599 1.72259800 -1.00867451
## V13 0.3869437 0.42586741 0.004607682 0.44065097 -0.96821289
## V14 0.5577645 0.58501403 -0.585943668 -0.26526148 -0.11783409
## V15 0.3826992 0.55820564 -0.581268835 0.04219305 -0.25048343
## V16 -0.3657428 -0.17931604 -0.599073658 -0.52354159 -0.73511476
## V17 0.8623497 0.62603976 -1.026463536 0.50129288 -0.52029859
## V18 0.9965917 1.43672898 -0.374120068 0.84003083 -0.48011513
## V19 0.6895183 0.86488105 -1.019763377 0.51455056 -0.08336993
## V20 1.0709497 1.12536111 -0.977848158 0.79229271 0.35427353
## V21 -0.3967921 -0.26546065 -0.865637295 1.24222863 -0.81031003
## V22 0.3541909 0.16391820 -0.981398436 1.10363674 -0.85610711
## V23 0.6809545 0.74906259 0.121361537 0.09421150 -0.16789856
## V24 0.8626260 1.03403338 -0.656405796 0.93112208 0.47421865
## V25 1.6918572 1.62156687 -0.485707157 1.11094938 0.97277085
## col. hue OD ratio proline
## V1 1.01813076 -0.5094731 -0.80114694
## V2 -0.23811536 -0.9758244 -0.46889055
## V3 -0.98400675 -1.2792975 -0.49166722
## V4 -1.48157185 -1.3751346 -0.23516832
## V5 -1.42975789 -1.4767622 -0.47276563
## V6 1.19383740 0.1846570 -0.67418566
## V7 -0.28471349 0.4891775 -1.08203396
## V8 -0.59809238 -0.2894469 -0.78569676
## V9 -1.07332524 -1.2175027 -0.21629667
## V10 -0.77347593 -0.9090949 0.09592919
## V11 -0.09058445 0.6891691 -0.91223551
## V12 0.03668272 0.8191445 -0.92841099
## V13 -0.17682497 0.7227036 -0.92435213
## V14 -0.28306939 0.7187163 0.72886711
## V15 0.52374694 0.8124692 1.27144941
## V16 0.95630863 0.1142933 -0.83567869
## V17 0.52414485 0.3870566 -0.44971048
## V18 0.50574815 1.0155404 -0.64590056
## V19 0.18104446 1.2421036 0.31367282
## V20 0.11033775 1.0290767 0.74444020
## V21 0.90568620 0.1585810 -0.23826560
## V22 0.78248753 0.3096531 0.24043601
## V23 0.73619680 0.6261033 0.76032841
## V24 0.74567202 0.5116907 1.71449830
## V25 0.53976730 0.5479133 2.04703252
peta<-cutree(hclust(dist(som.wines$codes[[1]])),5)
plot(peta)
plot(som.wines,type="codes",bgcol=rainbow(5)[peta])
add.cluster.boundaries(som.wines,peta)
