Global Financial Development Database PCA
2024-12-23
Installing packages and loading datset
#install.packages("corrplot")
library(corrplot)
#install.packages("factoextra")
library(factoextra)
library(readxl)
glob_financial_dev <- read_excel("C:/Users/stran/Desktop/UW/UL/glob_financial_dev.xlsx", col_types = c("text", "text", "numeric",
"text", "text", "text", "numeric",
"numeric", "numeric", "numeric",
"numeric", "numeric", "numeric",
"numeric", "numeric", "numeric",
"numeric", "numeric"))Checkiing correlation
cor_matrix <- cor(glob_financial_dev[,8:18], use = "complete.obs", method = "pearson")
corrplot(cor_matrix, method = "color", type = "upper", tl.cex = 0.8, tl.col = "black")
Filling NA’s
glob_financial_dev$ai10[is.na(glob_financial_dev$ai10)] <- mean((glob_financial_dev$ai10), na.rm = TRUE)
glob_financial_dev$ai23[is.na(glob_financial_dev$ai23)] <- mean((glob_financial_dev$ai23), na.rm = TRUE)Normalizing data
scaled_data <- scale(glob_financial_dev[,8:18])PCA & doing loadings
pca_result2 <- prcomp(scaled_data, center = FALSE, scale. = FALSE)summary(pca_result2)## Importance of components:
## PC1 PC2 PC3 PC4 PC5 PC6 PC7
## Standard deviation 2.6729 1.1866 1.02295 0.78589 0.55262 0.51620 0.30436
## Proportion of Variance 0.6495 0.1280 0.09513 0.05615 0.02776 0.02422 0.00842
## Cumulative Proportion 0.6495 0.7775 0.87263 0.92878 0.95654 0.98076 0.98919
## PC8 PC9 PC10 PC11
## Standard deviation 0.20849 0.18741 0.16438 0.11555
## Proportion of Variance 0.00395 0.00319 0.00246 0.00121
## Cumulative Proportion 0.99314 0.99633 0.99879 1.00000pca_result1 <- princomp(scaled_data)
loadings(pca_result1)##
## Loadings:
## Comp.1 Comp.2 Comp.3 Comp.4 Comp.5 Comp.6 Comp.7 Comp.8 Comp.9 Comp.10
## ai05 0.339 0.104 0.179 0.182 0.485 0.252 0.211 0.562 0.313
## ai06 0.352 -0.267 -0.268 0.613 -0.378 0.445 0.106
## ai07 0.347 -0.284 -0.200 0.313 0.182 0.291 -0.319
## ai10 -0.131 -0.576 0.539 -0.252 -0.117 0.465 0.207 -0.126
## ai12 0.302 -0.227 0.108 -0.530 -0.541 -0.219 0.338 -0.316
## ai14 0.195 -0.523 -0.566 -0.246
## ai18 -0.258 -0.495 -0.198 0.269 0.511 -0.184 -0.116 0.252
## ai20 0.351 -0.241 -0.197 0.278 -0.357 -0.474 0.586
## ai21 0.345 0.123 0.182 0.159 0.435 -0.395 -0.399 -0.550
## ai22 0.346 -0.139 0.231 0.136 -0.656 -0.185 0.531 0.115
## ai23 0.257 -0.223 0.286 0.669 -0.459 -0.347 0.112
## Comp.11
## ai05 0.227
## ai06
## ai07 -0.661
## ai10
## ai12
## ai14 0.531
## ai18 -0.443
## ai20
## ai21
## ai22 -0.147
## ai23
##
## Comp.1 Comp.2 Comp.3 Comp.4 Comp.5 Comp.6 Comp.7 Comp.8 Comp.9
## SS loadings 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
## Proportion Var 0.091 0.091 0.091 0.091 0.091 0.091 0.091 0.091 0.091
## Cumulative Var 0.091 0.182 0.273 0.364 0.455 0.545 0.636 0.727 0.818
## Comp.10 Comp.11
## SS loadings 1.000 1.000
## Proportion Var 0.091 0.091
## Cumulative Var 0.909 1.000Plots
plot(pca_result1)
fviz_pca_var(pca_result1, col.var="steelblue")
fviz_eig(pca_result1, choice='eigenvalue')
ai18 and ai10 are negetivly correlated with almost everything else.
ai18- The percentage of respondents who report borrowing any money from family, relatives, or friends in the past year
ai10- The percentage of respondents who report personally receiving any money in the past year from a relative or friend living in a different area of their country, and received it into a financial institution account or a mobile money account
ai07 - The percentage of respondents who report borrowing any money from a bank or another type of financial institution or using a credit card in the past year
Conclusions that i can extract:
If person borrows money from bank, they don’t tend to borrow it from relatives/friends since ai18 and ai07 are negatively correlated.
Person tends to borrow money from institutions or using credit card, when they have credit or debit card (ai20 and ai21)
3 first PC-s can big fraction of variance and over 85% of Cumulative Proportion