library(readxl)
corredores<-read_excel("C:/Users/MINEDUCYT/Desktop/corredores.xlsx")
corredores
# A tibble: 12 × 4
`km 4` `km 8` `km 12` `km 16`
<dbl> <dbl> <dbl> <dbl>
1 10 10 13 12
2 12 12 14 15
3 11 10 14 13
4 9 9 11 11
5 8 8 9 8
6 8 9 10 9
7 10 10 8 9
8 11 12 10 9
9 14 13 11 11
10 12 12 12 10
11 13 13 11 11
12 14 15 14 13
library(FactoMineR)
componentes <- PCA(corredores, graph = FALSE)
componentes$eig
eigenvalue percentage of variance cumulative percentage of variance
comp 1 2.88618524 72.154631 72.15463
comp 2 0.96729411 24.182353 96.33698
comp 3 0.10217301 2.554325 98.89131
comp 4 0.04434764 1.108691 100.00000
Calculando las componentes.
coef_PC1 <- componentes$var$coord[, 1]
coef_PC1
km 4 km 8 km 12 km 16
0.8776220 0.8407308 0.8342672 0.8444731
PC1 <- as.matrix(corredores)%*%coef_PC1
PC1
[,1]
[1,] 38.16268
[2,] 44.96707
[3,] 40.71904
[4,] 33.93132
[5,] 28.01101
[6,] 30.53048
[7,] 31.45792
[8,] 35.68554
[9,] 41.68235
[10,] 39.07617
[11,] 40.80473
[12,] 47.55556
coef_PC2 <- componentes$var$coord[, 2]
coef_PC2
km 4 km 8 km 12 km 16
-0.4521406 -0.5193155 0.5049832 0.4880229
PC2 <- as.matrix(corredores)%*%coef_PC2
PC2
[,1]
[1,] 2.7064950
[2,] 2.7326345
[3,] 3.2473605
[4,] 2.1799618
[5,] 0.6773829
[6,] 1.1510735
[7,] -1.2824897
[8,] -1.7632949
[9,] -2.1580033
[10,] -0.7174462
[11,] -1.7058627
[12,] -0.7056390
coef_PC3<-componentes$var$coord[,3]
coef_PC3
km 4 km 8 km 12 km 16
-0.06246401 0.06458405 0.21853379 -0.21527457
coef_PC4<-componentes$var$coord[,4]
coef_PC4
km 4 km 8 km 12 km 16
-0.14644697 0.13896759 -0.03511663 0.04853608
var_corredores<-var(corredores)
corr_corredores<-cor(corredores)
egenval<-eigen(corr_corredores)
egenval
eigen() decomposition
$values
[1] 2.88618524 0.96729411 0.10217301 0.04434764
$vectors
[,1] [,2] [,3] [,4]
[1,] -0.5165893 0.4597209 0.1954167 0.6954168
[2,] -0.4948743 0.5280220 -0.2020492 -0.6599002
[3,] -0.4910696 -0.5134494 -0.6836763 0.1667545
[4,] -0.4970771 -0.4962047 0.6734799 -0.2304780
var_explicada<-100*(sum(egenval$values[1:2])/sum(egenval$values[1:4]))
var_explicada
pca1 <- prcomp(corredores, scale = T)
pca1
Standard deviations (1, .., p=4):
[1] 1.6988776 0.9835111 0.3196451 0.2105888
Rotation (n x k) = (4 x 4):
PC1 PC2 PC3 PC4
km 4 0.5165893 0.4597209 0.1954167 0.6954168
km 8 0.4948743 0.5280220 -0.2020492 -0.6599002
km 12 0.4910696 -0.5134494 -0.6836763 0.1667545
km 16 0.4970771 -0.4962047 0.6734799 -0.2304780
Importance of components:
PC1 PC2 PC3 PC4
Standard deviation 1.6989 0.9835 0.31965 0.21059
Proportion of Variance 0.7216 0.2418 0.02554 0.01109
Cumulative Proportion 0.7216 0.9634 0.98891 1.00000
PC1 PC2
km 4 0.5165893 0.4597209
km 8 0.4948743 0.5280220
km 12 0.4910696 -0.5134494
km 16 0.4970771 -0.4962047
biplot(pca1, cex = c(0.01, 1), scale = 0.2, xlim=c(-2,2),ylim = c(-2, 2))
points(x = pca1$x[,1], y = pca1$x[,2], cex = 1, pch=19,col = "blue")
