Segmentación de Países por Clustering

1import pandas as pd
2import numpy as np
3import matplotlib.pyplot as plt
4import seaborn as sns
5
6df_country = pd.read_csv('/content/Country-data.csv')
7
8# Verificamos si hay valores nulos en el dataset
9df_country.isnull().sum()  # country       0
10                           # child_mort    0
11                           # exports       0
12                           # health        0
13                           # imports       0
14                           # income        0
15                           # inflation     0
16                           # life_expec    0
17                           # total_fer     0
18                           # gdpp          0

1df_country[df_country.duplicated()]
2
3int_cols = df_country.select_dtypes(exclude='object').columns
4for _ in int_cols:
5    plt.figure(figsize=(10,10))
6    sns.boxplot(data=df_country, y=_)
7    plt.show()

1fig = plt.figure(figsize=(15,10))
2sns.heatmap(df_country.corr(), annot=True, cmap='coolwarm')
3plt.show()

1from sklearn.preprocessing import StandardScaler
2scaaler = StandardScaler()
3df_country_scaled = scaaler.fit_transform(df_country.drop('country',axis=1))
4df_country_scaled = pd.DataFrame(df_country_scaled, columns = df_country.drop('country',axis=1).columns)
5
6from sklearn.decomposition import PCA
7pca = PCA()
8pca.fit(df_country_scaled)
9pca_data_scaled = pca.transform(df_country_scaled)
10
11var = pca.explained_variance_ratio_
12print(var)  # [0.4595174  0.17181626 0.13004259 0.11053162 0.07340211 0.02484235  0.0126043  0.00981282 0.00743056]
13
14cum_var = np.cumsum(np.round(var, decimals=4)*100)
15plt.figure(figsize=(10,10))
16plt.plot(cum_var, 'r-x')
17plt.show()

1pca_data_standard = pd.DataFrame(pca_data_scaled)
2pca_data_standard.drop([4,5,6,7,8], axis=1, inplace=True)
3
4# Ejemplo de cómo se ve el DataFrame resultante tras conservar solo las primeras 4 componentes principales:
5#           0         1         2         3
6# 0 -2.913025  0.095621 -0.718118  1.005255
7# 1  0.429911 -0.588156 -0.333486 -1.161059
8# 2 -0.285225 -0.455174  1.221505 -0.868115
9# 3 -2.932423  1.695555  1.525044  0.839625
10# 4  1.033576  0.136659 -0.225721 -0.847063