Entrega: clusters de tuestes de café
Julián Estivent Serna Triana
2/6/2022
Clustering
Cargando datos y haciendo data frame
dato <- read_excel("C:/Users/admin/Downloads/Cielab_tueste_cafe.xlsx")
datox <- data.frame(dato)
datoslab <- select(datox, -tueste)
#View(Cielab_tueste_cafe)Agregando Chroma y Hue
Chroma <- sqrt(datox$a^2 + datox$b^2)
Hue <- atan(datox$b/datox$a)
datos <- cbind(datox, Chroma, Hue)
datos <- select(datos, -a,-b)fig<- plot_ly(data = datos,
x = ~l,
y = ~Chroma,
z = ~Hue,
size = 0.7,
color = ~tueste)
fig## No trace type specified:
## Based on info supplied, a 'scatter3d' trace seems appropriate.
## Read more about this trace type -> https://plotly.com/r/reference/#scatter3d## No scatter3d mode specifed:
## Setting the mode to markers
## Read more about this attribute -> https://plotly.com/r/reference/#scatter-modeClustering
## Número óptimo de clusters
M = datos[ ,-2]
Ms = scale(M)
fviz_nbclust(Ms,
FUNcluster = kmeans,
method = 'gap_stat',
diss = get_dist(Ms,
'euclidean'))
## Caracterizando datos
clus = kmeans(Ms, 3)
datos$cluster <- clus$cluster
datosf <- datos[-2]
datosf |>
group_by(cluster) |>
summarise(media_Chroma = mean(Chroma),
media_Hue = mean(Hue),
media_l = mean(l),
desv_Chroma = sd(Chroma),
desv_Hue = sd(Hue),
desv_l = sd(l),
coeV_Chroma = 100 * desv_Chroma/media_Chroma,
coeV_Hue = 100 *desv_Hue/media_Hue,
coeV_l = 100 *desv_l/media_l)## # A tibble: 3 x 10
## cluster media_Chroma media_Hue media_l desv_Chroma desv_Hue desv_l coeV_Chroma
## <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 1 38.4 0.842 28.8 1.94 0.0153 1.55 5.05
## 2 2 25.3 0.823 16.9 3.43 0.0192 2.02 13.6
## 3 3 27.2 0.768 15.9 4.12 0.0208 2.71 15.2
## # ... with 2 more variables: coeV_Hue <dbl>, coeV_l <dbl>Gráficando clusters
# Dimension reduction using PCA
res.pca <- prcomp(datos[, -2], scale = TRUE)
# Coordinates of individuals
ind.coord <- as.data.frame(get_pca_ind(res.pca)$coord)
# Add clusters obtained using the K-means algorithm
ind.coord$cluster <- factor(clus$cluster)
# Add Species groups from the original data sett
ind.coord$tueste <- datos$tueste
# Data inspection
#head(ind.coord)eigenvalue <- round(get_eigenvalue(res.pca), 1)
variance.percent <- eigenvalue$variance.percent
#head(eigenvalue)
ggscatter(
ind.coord, x = "Dim.1", y = "Dim.2", title = "Clusters",
color = "cluster", palette = "npg", ellipse = TRUE, ellipse.type = "convex",
shape = "tueste", size = 1.5, legend = "right", ggtheme = theme_bw(),
xlab = paste0("Dim 1 (", variance.percent[1], "% )" ),
ylab = paste0("Dim 2 (", variance.percent[2], "% )" )
) +
stat_mean(aes(color = cluster), size = 4)
\[-El\ grupo\ 2\ representa\ el\ tostion\
claro\ de\ acuerdo\ a\ Luz,\ hue\ y\ Chroma \\ - Los\ tostiones\
medio,\ oscuro\ y\ verde\ se\ encuentran\ en\ el\ grupo\ 2\ y\ 3 \\ -
El\ tostión\ verde\ está\ en\ mayor\ proporción\ en\ el\ grupo\ 1 \\
\small Conclusión:\ Hay\ influencia\ del\ hue\ y\ el\ chroma\ en\ la\
conformación\ de\ clusters\ ya\ que\ a\ comparación\ con\ los\ grupos\
lab,\\ \small el\ tueste\ verde\ pasó\ a\ estar\ agrupado\ con\ los\
tuestes\ medio\ y\ oscuro\]
## Método de Ward
# Ward Hierarchical Clustering
df_scale <- scale(datosf)
d <- dist(df_scale, method = "euclidean") # distance matrix
fit <- hclust(d, method="ward.D2")
plot(fit) # display dendogram
groups <- cutree(fit, k=3) # cut tree into 5 clusters
# draw dendogram with red borders around the 5 clusters
rect.hclust(fit, k=3, border="red")
tapply(datosf$Chroma, groups, mean)## 1 2 3
## 27.16622 25.26230 38.37024tapply(datosf$Hue, groups, mean)## 1 2 3
## 0.7679318 0.8227090 0.8415492tapply(datosf$l, groups, mean)## 1 2 3
## 15.85863 16.90728 28.84291\[ Las\ medias\ coinciden\ con\ las\ del\ método\ Kmeans \]
Conversión a rgb
matriz <- as.matrix(datoslab)
rgbd <- convert_colour(matriz, "lab", "rgb")
rgbd <- data.frame(rgbd)
tueste <- datox$tueste
rgbdata <- cbind(rgbd, tueste)
rgbdata## r g b tueste
## 1 72.67764 20.92596 2.73267573 verde
## 2 62.80198 17.51391 0.00000000 verde
## 3 68.68871 17.37661 0.00000000 verde
## 4 65.91809 20.08825 0.79302516 verde
## 5 63.83378 20.99274 0.00000000 verde
## 6 63.92127 16.40077 0.00000000 verde
## 7 76.81804 22.85914 1.39417457 verde
## 8 68.36776 20.70484 0.00000000 verde
## 9 60.39985 17.12378 0.00000000 verde
## 10 70.80204 20.28769 0.07709687 verde
## 11 68.41939 16.22298 0.00000000 verde
## 12 71.29125 20.75246 0.00000000 verde
## 13 70.98326 20.73070 0.00000000 verde
## 14 67.93681 22.37354 0.28745091 verde
## 15 70.76505 21.49941 1.96940393 verde
## 16 69.03642 17.78844 0.00000000 verde
## 17 65.80367 17.11087 0.00000000 verde
## 18 63.92970 18.29821 0.00000000 verde
## 19 67.37263 19.83552 0.00000000 verde
## 20 68.77106 20.55613 0.00000000 verde
## 21 68.77614 21.98758 0.00000000 verde
## 22 68.74562 23.88500 0.66161441 verde
## 23 70.06856 18.07190 0.00000000 verde
## 24 60.24529 18.76858 1.01975592 verde
## 25 69.90397 19.83317 0.00000000 verde
## 26 63.94644 19.29920 1.85133011 verde
## 27 73.95803 23.00820 2.15978757 verde
## 28 56.36678 14.91563 0.00000000 verde
## 29 54.18067 15.58712 0.00000000 verde
## 30 66.11327 16.28020 0.00000000 verde
## 31 106.87410 49.58532 23.16851804 claro
## 32 99.66334 38.44458 16.02761489 claro
## 33 108.39261 48.66785 21.92270089 claro
## 34 121.20785 51.23165 23.59925178 claro
## 35 103.85262 45.55531 21.39241939 claro
## 36 118.66186 52.25492 25.98962496 claro
## 37 113.77134 47.52208 22.82939697 claro
## 38 116.74964 52.11466 23.68627677 claro
## 39 117.86707 53.88962 27.00060190 claro
## 40 117.97379 51.94398 24.41957047 claro
## 41 103.78617 46.52540 21.64925140 claro
## 42 123.55288 55.26656 26.13137881 claro
## 43 114.39631 50.90986 24.77053469 claro
## 44 112.80462 50.87835 24.61266124 claro
## 45 106.72050 51.42888 27.39301640 claro
## 46 111.32003 50.00690 24.13744731 claro
## 47 110.29886 49.45873 23.40996336 claro
## 48 111.10117 47.94954 22.93801277 claro
## 49 108.40412 52.18133 26.93807366 claro
## 50 114.32981 49.91905 24.04580901 claro
## 51 113.66298 50.03531 25.91257888 claro
## 52 111.41798 49.06226 24.62848560 claro
## 53 109.30148 47.34455 23.11818144 claro
## 54 104.37160 46.91030 23.20142866 claro
## 55 112.99048 54.24933 28.10964800 claro
## 56 111.01404 49.28273 23.72082811 claro
## 57 108.41972 48.73494 24.00020677 claro
## 58 113.44339 52.26206 28.54764446 claro
## 59 111.69406 51.31429 26.34352683 claro
## 60 118.69793 51.61667 23.91064680 claro
## 61 59.19814 27.32729 17.59884964 medio
## 62 65.80884 30.53196 15.13057211 medio
## 63 68.59459 30.80435 16.05367995 medio
## 64 66.35272 32.86064 17.48510704 medio
## 65 64.18658 28.78635 17.87586420 medio
## 66 65.27337 29.98946 19.95420083 medio
## 67 67.47122 30.42804 18.85395779 medio
## 68 71.76801 33.06048 18.05216165 medio
## 69 67.13516 32.38932 18.78026538 medio
## 70 83.34513 35.30951 20.25202425 medio
## 71 64.33502 32.94329 20.79545176 medio
## 72 68.73709 32.95713 21.74966318 medio
## 73 60.78390 28.28824 17.84776104 medio
## 74 69.75101 32.20084 18.82777130 medio
## 75 77.26769 32.63654 19.03741349 medio
## 76 66.14862 26.77646 10.50587565 medio
## 77 71.45927 31.11504 16.08009105 medio
## 78 67.31535 32.80723 18.81152109 medio
## 79 59.75839 30.05384 15.87061671 medio
## 80 68.95957 31.50741 18.60522686 medio
## 81 64.59914 30.40266 18.33234091 medio
## 82 60.21391 26.06094 14.99721535 medio
## 83 68.71319 32.02094 21.51584748 medio
## 84 70.52051 31.66818 17.40027451 medio
## 85 77.51342 35.42468 23.71882259 medio
## 86 69.28028 31.57800 19.23674232 medio
## 87 60.49671 29.31402 16.64622363 medio
## 88 66.77907 30.76577 21.30820563 medio
## 89 74.96630 35.78173 20.51951416 medio
## 90 62.07350 30.37804 17.52859297 medio
## 91 77.83784 32.84928 17.20769899 oscuro
## 92 70.97274 31.95185 18.67594824 oscuro
## 93 64.70535 29.03691 17.42533255 oscuro
## 94 66.51603 30.51775 17.31870484 oscuro
## 95 77.72817 32.10519 17.35750013 oscuro
## 96 68.79068 31.89727 20.90211736 oscuro
## 97 74.81262 36.52343 20.76098892 oscuro
## 98 75.73127 35.22184 22.32960723 oscuro
## 99 73.01286 35.36284 21.75627793 oscuro
## 100 73.68432 34.95950 21.69207126 oscuro
## 101 74.36798 29.93017 16.55624451 oscuro
## 102 73.52240 32.86448 20.57650578 oscuro
## 103 78.49626 35.92212 22.47465071 oscuro
## 104 76.72987 33.04649 22.48265074 oscuro
## 105 79.60777 35.05733 21.44949264 oscuro
## 106 76.59288 36.44203 22.29225342 oscuro
## 107 78.48510 36.48283 19.39674118 oscuro
## 108 71.26705 30.98873 14.26218059 oscuro
## 109 72.64089 33.59417 17.95779029 oscuro
## 110 78.00722 37.91390 23.31575167 oscuro
## 111 65.00828 28.75483 14.93933244 oscuro
## 112 71.53906 35.16540 25.09611230 oscuro
## 113 87.20987 40.68375 23.53446426 oscuro
## 114 66.81742 31.51330 16.95757930 oscuro
## 115 72.59407 33.22025 16.93933807 oscuro
## 116 70.11708 31.62267 17.07399406 oscuro
## 117 74.04029 31.59935 17.28935300 oscuro
## 118 71.08770 33.57784 20.62737112 oscuro
## 119 83.02760 36.60495 21.21118220 oscuro
## 120 70.31887 35.43990 21.82361742 oscurograficando
figrgb <- plot_ly(data = rgbdata,
x = ~r,
y = ~g,
z = ~b,
size = 0.7,
color = ~tueste)
figrgb## No trace type specified:
## Based on info supplied, a 'scatter3d' trace seems appropriate.
## Read more about this trace type -> https://plotly.com/r/reference/#scatter3d## No scatter3d mode specifed:
## Setting the mode to markers
## Read more about this attribute -> https://plotly.com/r/reference/#scatter-mode`` ## Clustering
## Número óptimo de clusters
M1 = rgbdata[ ,-4]
Ms1 = scale(M1)
fviz_nbclust(Ms1,
FUNcluster = kmeans,
method = 'gap_stat',
diss = get_dist(Ms1,
'euclidean'))
## Caracterizando datos
clus1 = kmeans(Ms1, 3)
rgbdata$cluster <- clus1$cluster
rgbdataf <- rgbdata[-4]
rgbdataf |>
group_by(cluster) |>
summarise(media_r = mean(r),
media_g = mean(g),
media_b = mean(b),
desv_r = sd(r),
desv_g = sd(g),
desv_b = sd(b),
coeV_r = 100 *desv_r/media_r,
coeV_g = 100 *desv_g/media_g,
coeV_b = 100 *desv_b/media_b)## # A tibble: 3 x 10
## cluster media_r media_g media_b desv_r desv_g desv_b coeV_r coeV_g coeV_b
## <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 1 70.7 32.5 19.0 6.16 2.83 2.73 8.72 8.72 14.3
## 2 2 67.0 19.4 0.432 4.93 2.39 0.788 7.36 12.3 183.
## 3 3 112. 49.9 24.3 5.48 3.17 2.42 4.90 6.35 9.97Gráficando clusters
# Dimension reduction using PCA
res.pca <- prcomp(rgbdata[, -4], scale = TRUE)
# Coordinates of individuals
ind.coord <- as.data.frame(get_pca_ind(res.pca)$coord)
# Add clusters obtained using the K-means algorithm
ind.coord$cluster <- factor(clus1$cluster)
# Add Species groups from the original data sett
ind.coord$tueste <- rgbdata$tueste
# Data inspection
#head(ind.coord)eigenvalue <- round(get_eigenvalue(res.pca), 1)
variance.percent <- eigenvalue$variance.percent
#head(eigenvalue)
ggscatter(
ind.coord, x = "Dim.1", y = "Dim.2", title = "Clusters rgb",
color = "cluster", palette = "npg", ellipse = TRUE, ellipse.type = "convex",
shape = "tueste", size = 1.5, legend = "right", ggtheme = theme_bw(),
xlab = paste0("Dim 1 (", variance.percent[1], "% )" ),
ylab = paste0("Dim 2 (", variance.percent[2], "% )" )
) +
stat_mean(aes(color = cluster), size = 4)
\[-El\ grupo\ 1\ representa\ ambos\ tuestes\
medio\ y\ oscuro\ de\ acuerdo\ a\ rgb \\ - El\ tueste\ claro,\
pertenece\ a\ el\ grupo\ 3 \\ - El\ tueste\ verde\ está\ en\ el\
grupo\ 2 \\ \small Conclusión:\ tanto\ en\ lab\ como\ en\ rgb\ los\
tuestes\ verdes\ y\ claros\ están\ en\ grupos\ separados\ mientras\ que\
los\\ tuestes\ medio\ y\ oscuro\ comparten\ grupo\]
## Método de Ward
# Ward Hierarchical Clustering
df1_scale <- scale(rgbdataf)
d1 <- dist(df1_scale, method = "euclidean") # distance matrix
fit1 <- hclust(d1, method="ward.D2")
plot(fit1) # display dendogram
groups <- cutree(fit1, k=3) # cut tree into 5 clusters
# draw dendogram with red borders around the 5 clusters
rect.hclust(fit, k=3, border="red")
tapply(rgbdataf$r, groups, mean)## 1 2 3
## 67.02811 111.89141 70.73459tapply(rgbdata$g, groups, mean)## 1 2 3
## 19.36928 49.88490 32.45031tapply(rgbdata$b, groups, mean)## 1 2 3
## 0.4315438 24.2518433 19.0175785\[Las\ medias\ coinciden\ con\ el\ método\ Kmeans\]
hsvd <- convert_colour(matriz, "lab", "hsv")
hsvd <- data.frame(hsvd)
hsvdata <- cbind(hsvd, tueste)
hsvdata## h s v tueste
## 1 15.60651 0.9624000 0.2850104 verde
## 2 20.00781 1.0000000 0.2462823 verde
## 3 19.68317 1.0000000 0.2693675 verde
## 4 17.77677 0.9879695 0.2585023 verde
## 5 20.30000 1.0000000 0.2503286 verde
## 6 18.34603 1.0000000 0.2506717 verde
## 7 17.07547 0.9818509 0.3012472 verde
## 8 21.44449 1.0000000 0.2681088 verde
## 9 19.23513 1.0000000 0.2368622 verde
## 10 17.14580 0.9989111 0.2776551 verde
## 11 18.93798 1.0000000 0.2683113 verde
## 12 18.48172 1.0000000 0.2795735 verde
## 13 19.07124 1.0000000 0.2783657 verde
## 14 19.58873 0.9957688 0.2664188 verde
## 15 17.03306 0.9721698 0.2775100 verde
## 16 18.70100 1.0000000 0.2707310 verde
## 17 18.50810 1.0000000 0.2580536 verde
## 18 19.16756 1.0000000 0.2507047 verde
## 19 18.82095 1.0000000 0.2642064 verde
## 20 20.90034 1.0000000 0.2696904 verde
## 21 19.83876 1.0000000 0.2697104 verde
## 22 20.46594 0.9903759 0.2695907 verde
## 23 18.44496 1.0000000 0.2747787 verde
## 24 17.98091 0.9830733 0.2362560 verde
## 25 19.70933 1.0000000 0.2741332 verde
## 26 16.85917 0.9710487 0.2507703 verde
## 27 17.42250 0.9707971 0.2900315 verde
## 28 21.90418 1.0000000 0.2210462 verde
## 29 21.49104 1.0000000 0.2124732 verde
## 30 18.25407 1.0000000 0.2592677 verde
## 31 18.93551 0.7832167 0.4191141 claro
## 32 16.08186 0.8391824 0.3908366 claro
## 33 18.55800 0.7977473 0.4250690 claro
## 34 16.98563 0.8052993 0.4753249 claro
## 35 17.58150 0.7940117 0.4072652 claro
## 36 17.00528 0.7809774 0.4653406 claro
## 37 16.29128 0.7993397 0.4461621 claro
## 38 18.32841 0.7971191 0.4578417 claro
## 39 17.75507 0.7709233 0.4622238 claro
## 40 17.65249 0.7930085 0.4626423 claro
## 41 18.17172 0.7914052 0.4070046 claro
## 42 17.94379 0.7885005 0.4845211 claro
## 43 17.49898 0.7834674 0.4486130 claro
## 44 17.86944 0.7818116 0.4423711 claro
## 45 18.17972 0.7433200 0.4185118 claro
## 46 17.80364 0.7831707 0.4365492 claro
## 47 17.98764 0.7877588 0.4325446 claro
## 48 17.02175 0.7935394 0.4356909 claro
## 49 18.59174 0.7515032 0.4251142 claro
## 50 17.19457 0.7896803 0.4483522 claro
## 51 16.49410 0.7720227 0.4457372 claro
## 52 16.89175 0.7789541 0.4369333 claro
## 53 16.86617 0.7884916 0.4286332 claro
## 54 17.52531 0.7777036 0.4093004 claro
## 55 18.47744 0.7512211 0.4430999 claro
## 56 17.56968 0.7863259 0.4353492 claro
## 57 17.57987 0.7786362 0.4251754 claro
## 58 16.76014 0.7483534 0.4448760 claro
## 59 17.55403 0.7641457 0.4380159 claro
## 60 17.53781 0.7985589 0.4654821 claro
## 61 14.03165 0.7027128 0.2321496 medio
## 62 18.23431 0.7700830 0.2580739 medio
## 63 16.84479 0.7659629 0.2689984 medio
## 64 18.87819 0.7364824 0.2602068 medio
## 65 14.13559 0.7215016 0.2517121 medio
## 66 13.28610 0.6942980 0.2559740 medio
## 67 14.28391 0.7205629 0.2645930 medio
## 68 16.76412 0.7484651 0.2814432 medio
## 69 16.88647 0.7202619 0.2632751 medio
## 70 14.31930 0.7570101 0.3268437 medio
## 71 16.74041 0.6767631 0.2522942 medio
## 72 14.31123 0.6835818 0.2695572 medio
## 73 14.58978 0.7063735 0.2383682 medio
## 74 15.75674 0.7300717 0.2735334 medio
## 75 14.01243 0.7536174 0.3030106 medio
## 76 17.54470 0.8411777 0.2594064 medio
## 77 16.28946 0.7749754 0.2802324 medio
## 78 17.31291 0.7205463 0.2639818 medio
## 79 19.39021 0.7344203 0.2343466 medio
## 80 15.37367 0.7302009 0.2704297 medio
## 81 15.65310 0.7162138 0.2533300 medio
## 82 14.68094 0.7509344 0.2361330 medio
## 83 13.35468 0.6868746 0.2694635 medio
## 84 16.11578 0.7532594 0.2765510 medio
## 85 13.05617 0.6940037 0.3039742 medio
## 86 14.79662 0.7223345 0.2716874 medio
## 87 17.33317 0.7248408 0.2372420 medio
## 88 12.47951 0.6809149 0.2618787 medio
## 89 16.81887 0.7262835 0.2939855 medio
## 90 17.30763 0.7176155 0.2434255 medio
## 91 15.47902 0.7789289 0.3052464 oscuro
## 92 15.23141 0.7368574 0.2783245 oscuro
## 93 14.73550 0.7306972 0.2537465 oscuro
## 94 16.09727 0.7396311 0.2608472 oscuro
## 95 14.65714 0.7766897 0.3048163 oscuro
## 96 13.77592 0.6961490 0.2697674 oscuro
## 97 17.49709 0.7224935 0.2933828 oscuro
## 98 14.48520 0.7051468 0.2969854 oscuro
## 99 15.92759 0.7020213 0.2863249 oscuro
## 100 15.31085 0.7056080 0.2889581 oscuro
## 101 13.88015 0.7773740 0.2916391 oscuro
## 102 13.92513 0.7201328 0.2883231 oscuro
## 103 14.40245 0.7136851 0.3078285 oscuro
## 104 11.68411 0.7069896 0.3009014 oscuro
## 105 14.03876 0.7305603 0.3121873 oscuro
## 106 15.63494 0.7089514 0.3003643 oscuro
## 107 17.34970 0.7528608 0.3077847 oscuro
## 108 17.60539 0.7998769 0.2794786 oscuro
## 109 17.15672 0.7527868 0.2848662 oscuro
## 110 16.01510 0.7011078 0.3059107 oscuro
## 111 16.55577 0.7701934 0.2549344 oscuro
## 112 13.00859 0.6491971 0.2805453 oscuro
## 113 16.15941 0.7301399 0.3419995 oscuro
## 114 17.51596 0.7462102 0.2620291 oscuro
## 115 17.55205 0.7666567 0.2846826 oscuro
## 116 16.45682 0.7564931 0.2749690 oscuro
## 117 15.12926 0.7664872 0.2903541 oscuro
## 118 15.39880 0.7098321 0.2787753 oscuro
## 119 14.94144 0.7445285 0.3255984 oscuro
## 120 16.84654 0.6896478 0.2757603 oscuroAnálisis gráfico
figrgb2 <- plot_ly(data = hsvdata,
x = ~h,
y = ~s,
z = ~v,
size = 0.7,
color = ~tueste)
figrgb2## No trace type specified:
## Based on info supplied, a 'scatter3d' trace seems appropriate.
## Read more about this trace type -> https://plotly.com/r/reference/#scatter3d## No scatter3d mode specifed:
## Setting the mode to markers
## Read more about this attribute -> https://plotly.com/r/reference/#scatter-modeM2 = hsvdata[ ,-4]
Ms2 = scale(M2)
fviz_nbclust(Ms2,
FUNcluster = kmeans,
method = 'gap_stat',
diss = get_dist(Ms2,
'euclidean'))
clus2 = kmeans(Ms2, 3)
hsvdata$cluster <- clus2$cluster
hsvf <- hsvdata[-4]# Dimension reduction using PCA
res.pca <- prcomp(hsvdata[, -4], scale = TRUE)
# Coordinates of individuals
ind.coord <- as.data.frame(get_pca_ind(res.pca)$coord)
# Add clusters obtained using the K-means algorithm
ind.coord$cluster <- factor(clus2$cluster)
# Add Species groups from the original data sett
ind.coord$tueste <- hsvdata$tueste
# Data inspection
#head(ind.coord)eigenvalue <- round(get_eigenvalue(res.pca), 1)
variance.percent <- eigenvalue$variance.percent
#head(eigenvalue)
ggscatter(
ind.coord, x = "Dim.1", y = "Dim.2", title = "Clusters hsv",
color = "cluster", palette = "npg", ellipse = TRUE, ellipse.type = "convex",
shape = "tueste", size = 1.5, legend = "right", ggtheme = theme_bw(),
xlab = paste0("Dim 1 (", variance.percent[1], "% )" ),
ylab = paste0("Dim 2 (", variance.percent[2], "% )" )
) +
stat_mean(aes(color = cluster), size = 4)
\[ \small Conclusión:\ La\ conformación\ de\
grupos\ en\ hsv\ sigue\ siendo\ la\ misma,\ pero\ cambia\ la\ magnitud\
respecto\ a\ rgb\ dónde\\ \small los\ tamaños\ del\ grupo\ 1\ y\ 2\ son\
inversos\ a\ los\ observados\ en\ hsv.\\ \small En\ general\ habría\
que\ considerar\ los\ tuestes\ medio\ y\ oscuro\ como\ uno\ mismo\ en\
base\ a\ los\ resultados\ observados \]