# Importar las librerías.
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
import statsmodels.formula.api as smf
import statsmodels.api as sm
from sklearn.preprocessing import LabelEncoder
from statsmodels.stats.outliers_influence import variance_inflation_factor
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import ConfusionMatrixDisplay, classification_report
from sklearn.linear_model import LogisticRegression
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
from sklearn.tree import DecisionTreeClassifier as DTC
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score, f1_score
from sklearn.model_selection import cross_val_score

# Cargar los datos.
url = 'https://raw.githubusercontent.com/estefaniadelarosa/IA-I/refs/heads/main/P1.%20Regresi%C3%B3n/P1.%20Regresi%C3%B3n/2024_2025_salud_mental.csv'
df = pd.read_csv(url)
print(df.shape)
df.head()

(48224, 14)

df['municipio_unidad_medica'].unique()

array(['LINARES', 'MONTEMORELOS', 'MONTERREY', 'SABINAS HIDALGO',
       'SAN NICOLAS DE LOS GARZA', 'SANTA CATARINA', 'GENERAL ESCOBEDO',
       'PESQUERIA', 'CADEREYTA JIMENEZ', 'SAN PEDRO GARZA GARCIA',
       'GARCIA', 'GALEANA', 'CERRALVO', 'CHINA'], dtype=object)

df1 = df[df['municipio_unidad_medica'] == 'SAN PEDRO GARZA GARCIA']
print(df1.shape)
df1.head()

(1060, 14)

# Aplicar el filtro de personas dentro del rango de edad de 18 a 65 años.
df2 = df1[(df1['edad'] >= 18) & (df1['edad'] <= 65)]
print(df2.shape)
df2.head()

(720, 14)

# Verificar si hay datos vacíos.
df2.isna().sum().sort_values(ascending = False)

fecha                           0
id_consulta                     0
edad                            0
edad_meses                      0
edad_dias                       0
sexo                            0
peso                            0
altura                          0
municipio_unidad_medica         0
institucion_unidad_medica       0
clave_grupo_ enfermedad         0
descripcion_grupo_enfermedad    0
clave_enfermedad                0
descripcion_enfermedad          0
dtype: int64

df2['descripcion_grupo_enfermedad'].value_counts()

descripcion_grupo_enfermedad
TRASTORNOS MENTALES Y DEL COMPORTAMIENTO                                             547
FACTORES QUE INFLUYEN EN EL ESTADO DE SALUD Y CONTACTO CON LOS SERVICIOS DE SALUD    173
Name: count, dtype: int64

# Para el porcentaje.
df2['descripcion_grupo_enfermedad'].value_counts()/(df2['descripcion_grupo_enfermedad'].value_counts().sum())*100

descripcion_grupo_enfermedad
TRASTORNOS MENTALES Y DEL COMPORTAMIENTO                                             75.972222
FACTORES QUE INFLUYEN EN EL ESTADO DE SALUD Y CONTACTO CON LOS SERVICIOS DE SALUD    24.027778
Name: count, dtype: float64

# Gráfica del balance de clases.
df2['descripcion_grupo_enfermedad'].str[:43].value_counts().plot(kind = 'bar')
plt.show()

df2.dtypes

fecha                           object
id_consulta                     object
edad                             int64
edad_meses                       int64
edad_dias                        int64
sexo                            object
peso                            object
altura                          object
municipio_unidad_medica         object
institucion_unidad_medica       object
clave_grupo_ enfermedad         object
descripcion_grupo_enfermedad    object
clave_enfermedad                object
descripcion_enfermedad          object
dtype: object

# Convertimos las variables categóricas a categóricas numéricas.
from sklearn.preprocessing import LabelEncoder 

# Aplicamos la transformación.
df2['sexo_num'] = LabelEncoder().fit_transform(df2['sexo'])
df2['institucion_unidad_medica_num'] = LabelEncoder().fit_transform(df2['institucion_unidad_medica'])
df2['descripcion_grupo_enfermedad_num'] = LabelEncoder().fit_transform(df2['descripcion_grupo_enfermedad'])
df2['clave_enfermedad_num'] = LabelEncoder().fit_transform(df2['clave_enfermedad'])
df2['descripcion_enfermedad_num'] = LabelEncoder().fit_transform(df2['descripcion_enfermedad'])

df2.sample(5)

/var/folders/rw/krrrqrzn68j3jq_d4yl8mzk80000gn/T/ipykernel_57416/763801538.py:5: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df2['sexo_num'] = LabelEncoder().fit_transform(df2['sexo'])
/var/folders/rw/krrrqrzn68j3jq_d4yl8mzk80000gn/T/ipykernel_57416/763801538.py:6: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df2['institucion_unidad_medica_num'] = LabelEncoder().fit_transform(df2['institucion_unidad_medica'])
/var/folders/rw/krrrqrzn68j3jq_d4yl8mzk80000gn/T/ipykernel_57416/763801538.py:7: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df2['descripcion_grupo_enfermedad_num'] = LabelEncoder().fit_transform(df2['descripcion_grupo_enfermedad'])
/var/folders/rw/krrrqrzn68j3jq_d4yl8mzk80000gn/T/ipykernel_57416/763801538.py:8: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df2['clave_enfermedad_num'] = LabelEncoder().fit_transform(df2['clave_enfermedad'])
/var/folders/rw/krrrqrzn68j3jq_d4yl8mzk80000gn/T/ipykernel_57416/763801538.py:9: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df2['descripcion_enfermedad_num'] = LabelEncoder().fit_transform(df2['descripcion_enfermedad'])

df2 = df2.drop(columns=['edad_meses', 'edad_dias'])

# Convertir peso y altura a enteros.
df2['peso'] = df2['peso'].astype(float)
df2['altura'] = df2['altura'].astype(float)

# Verificar.
print(df2.dtypes[['peso','altura']])

peso      float64
altura    float64
dtype: object

# Convertir altura a metros.
df2['altura_m'] = df2['altura'] / 100
# IMC.
df2['IMC'] = df2['peso'] / (df2['altura_m']**2)

plt.figure(figsize = (10, 6))
sns.heatmap(df2.corr(numeric_only = True), annot = True)
plt.show()

# Definimos el modelo de Feature Selection - ANOVA.
modelFS = smf.ols(formula='descripcion_grupo_enfermedad_num ~ edad + sexo_num + IMC + institucion_unidad_medica_num + clave_enfermedad_num + descripcion_enfermedad_num', data = df2).fit()
modelFS.summary()

# Imprimimos los P value para ver cuál es más importante.
pvalue_percent = (modelFS.pvalues * 100).sort_values(ascending = True)
pvalue_percent

clave_enfermedad_num             6.584618e-56
Intercept                        6.892180e-50
descripcion_enfermedad_num       3.738917e-06
sexo_num                         3.829040e-06
institucion_unidad_medica_num    6.021948e+00
IMC                              3.815116e+01
edad                             5.652051e+01
dtype: float64

X = df2[['edad', 'sexo_num', 'IMC', 'institucion_unidad_medica_num', 'clave_enfermedad_num', 'descripcion_enfermedad_num']]

vif_data = pd.DataFrame()
vif_data["feature"] = X.columns

vif_data["VIF"] = [variance_inflation_factor(X.values, i)
                          for i in range(len(X.columns))]
print(vif_data)

                         feature        VIF
0                           edad   9.406207
1                       sexo_num   2.008371
2                            IMC  20.720012
3  institucion_unidad_medica_num   2.930642
4           clave_enfermedad_num   8.236488
5     descripcion_enfermedad_num   6.445571

X = df2[['edad', 'sexo_num', 'institucion_unidad_medica_num', 'descripcion_enfermedad_num']]

vif_data = pd.DataFrame()
vif_data["feature"] = X.columns

vif_data["VIF"] = [variance_inflation_factor(X.values, i)
                          for i in range(len(X.columns))]
print(vif_data)

                         feature       VIF
0                           edad  4.187639
1                       sexo_num  1.918244
2  institucion_unidad_medica_num  2.000134
3     descripcion_enfermedad_num  4.060178

# Definimos entrada y salida.
X = df2[['edad', 'sexo_num', 'institucion_unidad_medica_num']]
y = df2['descripcion_grupo_enfermedad_num']
print(X.shape)
print(y.shape)

(720, 3)
(720,)

# Dividir los datos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0, stratify = y)
X_train.shape, X_test.shape, y_train.shape, y_test.shape
# No dividir de manera que se desbalanceen los datos -> cuidado con el random_state).

((576, 3), (144, 3), (576,), (144,))

# Definimos el modelo.
model1 = LogisticRegression(random_state = 0)
# Entrenamos el modelo.
model1.fit(X_train, y_train)

LogisticRegression(random_state=0)

scores = cross_val_score(model1, X, y, cv = 5)
print(scores.mean())

0.775

OR = np.exp(model1.coef_[0])
coef_df = pd.DataFrame({
  "Variables": X.columns,
  "Coeficientes": model1.coef_[0],
  "Odds Ratio": OR
})

print(coef_df)

                       Variables  Coeficientes  Odds Ratio
0                           edad     -0.025575    0.974749
1                       sexo_num      0.536856    1.710620
2  institucion_unidad_medica_num     -1.831048    0.160246

# Validar si el modelo pronostica adecuadamente.
y_pred_test = model1.predict(X_test)
print(y_pred_test[0:5])
print(y_test.head())

[1 1 1 1 1]
46382    0
32275    0
7409     1
32248    1
46407    1
Name: descripcion_grupo_enfermedad_num, dtype: int64

y_proba = model1.predict_proba(X)[:, 1]
y_pred = (y_proba >= 0.8).astype(int)
print(y_pred[0:5])

[1 1 1 1 1]

accuracy_train = model1.score(X_train, y_train)
print('Accuracy train = {:.2f}'.format(accuracy_train))
accuracy_test = model1.score(X_test, y_test)
print('Accuracy test = {:.2f}'.format(accuracy_test))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train-accuracy_test)*100))

Accuracy train = 0.79
Accuracy test = 0.82
Diferencia = 3.2986%

predRF = model1.predict(X_test)
repDT = classification_report(y_test, predRF)
print(repDT)

from imblearn.metrics import geometric_mean_score
print('G-mean =', geometric_mean_score(y_test, predRF))

disp = ConfusionMatrixDisplay.from_predictions(y_test, predRF, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.76      0.37      0.50        35
           1       0.83      0.96      0.89       109

    accuracy                           0.82       144
   macro avg       0.80      0.67      0.69       144
weighted avg       0.81      0.82      0.80       144

G-mean = 0.5981623234019464

from IPython.display import Image
Image(filename='/Users/estefaniadelarosa/Downloads/roc-curve-v2.png')

from sklearn.model_selection import cross_val_predict
from sklearn.metrics import roc_curve, roc_auc_score

# Validación cruzada.
y_probs = cross_val_predict(model1, X, y, cv = 5, method = 'predict_proba')[:,1] # Agregando método como scores.

# Curva ROC.
fpr, tpr, thresholds = roc_curve(y, y_probs)
auc = roc_auc_score(y, y_probs)

# Graficar
plt.plot(fpr, tpr)
plt.plot([0,1], [0,1])
plt.xlabel("False Positive Rate.")
plt.ylabel("True Positive Rate.")
plt.title("ROC Curve.")
plt.show()

print("AUC:", auc)

AUC: 0.7687227229977491

from imblearn.over_sampling import SMOTE

# Definir la técnica.
smote = SMOTE(random_state = 0)
# Aplicamos.
X_smote, y_smote = smote.fit_resample(X, y)

# Comprobar si funicona.
print('Tamaño de X antes de SMOTE:', X.shape)
print('Tamaño de X después de SMOTE:', X_smote.shape)
print('Balance de clases con SMOTE:', y_smote.value_counts())
print('Nuestras clases están balanceadas.')

Tamaño de X antes de SMOTE: (720, 3)
Tamaño de X después de SMOTE: (1094, 3)
Balance de clases con SMOTE: descripcion_grupo_enfermedad_num
1    547
0    547
Name: count, dtype: int64
Nuestras clases están balanceadas.

# Dividir los tratos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X_smote, y_smote, test_size = 0.2, random_state = 0, stratify = y_smote)
print(y_train.value_counts())
X_train.shape, X_test.shape, y_train.shape, y_test.shape
# No dividir de manera que se desbalanceen los datos -> cuidado con el random_state).

descripcion_grupo_enfermedad_num
1    438
0    437
Name: count, dtype: int64

((875, 3), (219, 3), (875,), (219,))

# Definimos el modelo.
model2 = LogisticRegression(random_state = 0)
# Entrenamos el modelo.
model2.fit(X_train, y_train)

LogisticRegression(random_state=0)

scores = cross_val_score(model2, X_smote, y_smote, cv = 5)
print(scores.mean())

0.7194294332034686

OR = np.exp(model2.coef_[0])
coef_df = pd.DataFrame({
  "Variables": X.columns,
  "Coeficientes": model2.coef_[0],
  "Odds Ratio": OR
})

print(coef_df)

                       Variables  Coeficientes  Odds Ratio
0                           edad     -0.022775    0.977482
1                       sexo_num      1.005345    2.732851
2  institucion_unidad_medica_num     -1.639948    0.193990

# Validar si el modelo pronostica adecuadamente.
y_pred_test = model2.predict(X_test)
print(y_pred_test[0:5])
print(y_test.head())

[0 1 0 1 0]
944     0
199     1
1034    0
405     0
727     0
Name: descripcion_grupo_enfermedad_num, dtype: int64

y_proba = model2.predict_proba(X)[:, 1]
y_pred = (y_proba >= 0.8).astype(int)
print(y_pred[0:5])

[1 0 0 0 0]

accuracy_train = model2.score(X_train, y_train)
print('Accuracy train = {:.2f}'.format(accuracy_train))
accuracy_test = model2.score(X_test, y_test)
print('Accuracy test = {:.2f}'.format(accuracy_test))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train-accuracy_test)*100))

Accuracy train = 0.72
Accuracy test = 0.74
Diferencia = 2.6578%

predRF = model2.predict(X_test)
repDT = classification_report(y_test, predRF)
print(repDT)

from imblearn.metrics import geometric_mean_score
print('G-mean =', geometric_mean_score(y_test, predRF))

disp = ConfusionMatrixDisplay.from_predictions(y_test, predRF, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.72      0.80      0.76       110
           1       0.77      0.69      0.73       109

    accuracy                           0.74       219
   macro avg       0.75      0.74      0.74       219
weighted avg       0.75      0.74      0.74       219

G-mean = 0.7419290502442469

from sklearn.model_selection import cross_val_predict
from sklearn.metrics import roc_curve, roc_auc_score
import matplotlib.pyplot as plt

# Validación cruzada.
y_probs = cross_val_predict(model2, X, y, cv = 5, method = 'predict_proba')[:,1] # Agregando método como scores.

# Curva ROC.
fpr, tpr, thresholds = roc_curve(y, y_probs)
auc = roc_auc_score(y, y_probs)

# Graficar.
plt.plot(fpr, tpr)
plt.plot([0,1], [0,1])
plt.xlabel("False Positive Rate.")
plt.ylabel("True Positive Rate.")
plt.title("ROC Curve.")
plt.show()

print("AUC:", auc)

AUC: 0.7687227229977491

# Definimos entrada y salida.
X = df2[['edad', 'sexo_num', 'institucion_unidad_medica_num']]
y = df2['descripcion_grupo_enfermedad_num']
print(X.shape)
print(y.shape)

(720, 3)
(720,)

# Dividir los datos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0, stratify = y)
X_train.shape, X_test.shape, y_train.shape, y_test.shape
# No dividir de manera que se desbalanceen los datos -> cuidado con el random_state).

((576, 3), (144, 3), (576,), (144,))

# Create a pair plot to visualize relationships between different features and species.
ax = sns.pairplot(df2, hue = 'descripcion_grupo_enfermedad_num', markers = ["o", "s", "D"])
plt.suptitle("Pair Plot: Grupos de Enfermedades.")
sns.move_legend(
    ax, "lower center",
    bbox_to_anchor = (.5, 1), ncol = 3, title = None, frameon = False)
plt.tight_layout()
plt.show()

/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)
/opt/anaconda3/lib/python3.13/site-packages/seaborn/axisgrid.py:1615: UserWarning: The markers list has more values (3) than needed (2), which may not be intended.
  func(x=x, y=y, **kwargs)

# Visualize the distribution of each feature using histograms.
plt.figure(figsize =( 12, 6))
for i, feature in enumerate(X[:-1]):
    plt.subplot(2, 2, i + 1)
    sns.histplot(data = df2, x = feature, hue = 'descripcion_grupo_enfermedad_num', kde = True)
    plt.title(f'{feature} Distribution.')

plt.tight_layout()
plt.show()

correlation_matrix = df2.corr(numeric_only = True)
plt.figure(figsize=(8, 6))
sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', linewidths=0.5)
plt.title("Correlation Heatmap")
plt.show()

# Verificación de n_components = 1 siendo 2 clases.
print(np.unique(y_train))
print(len(np.unique(y_train)))

[0 1]
2

# Apply Linear Discriminant Analysis.
lda = LinearDiscriminantAnalysis(n_components = 1)
X_train_lda = lda.fit_transform(X_train, y_train)
X_test_lda = lda.transform(X_test)

tmp_Df = pd.DataFrame(X_train_lda, columns = ['LDA Component 1'])
tmp_Df['Class'] = y_train.values  

# Grafica.
sns.histplot(data = tmp_Df, x = 'LDA Component 1', hue = 'Class', kde = True)
plt.show()

print(tmp_Df.head())

   LDA Component 1  Class
0        -0.695258      0
1        -1.551494      0
2         0.251022      1
3        -1.727778      1
4         0.580071      1

# Dividir los datos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0, stratify = y)
X_train.shape, X_test.shape, y_train.shape, y_test.shape
# No dividir de manera que se desbalanceen los datos -> cuidado con el random_state).

((576, 3), (144, 3), (576,), (144,))

from imblearn.over_sampling import SMOTE
smote = SMOTE(random_state = 0)
X_train_smote, y_train_smote = smote.fit_resample(X_train, y_train)

# Apply Linear Discriminant Analysis.
lda = LinearDiscriminantAnalysis(n_components = 1)
X_train_lda = lda.fit_transform(X_train_smote, y_train_smote)
X_test_lda = lda.transform(X_test)

tmp_Df = pd.DataFrame(X_train_lda, columns = ['LDA Component 1'])
tmp_Df['Class'] = y_train_smote.values  

# Grafica.
sns.histplot(data = tmp_Df, x = 'LDA Component 1', hue = 'Class', kde = True)
plt.show()

print(tmp_Df.head())

   LDA Component 1  Class
0        -0.571684      0
1        -1.299334      0
2         0.177260      1
3        -1.449144      1
4         1.296899      1

print(tmp_Df.head())

   LDA Component 1  Class
0        -0.571684      0
1        -1.299334      0
2         0.177260      1
3        -1.449144      1
4         1.296899      1

tree = DTC().fit(X_train, y_train)

from sklearn.tree import plot_tree
plt.figure(figsize=(15,10))
plot_tree(tree, filled = True, feature_names = X_train.columns)

[Text(0.4519230769230769, 0.9615384615384616, 'institucion_unidad_medica_num <= 0.5\ngini = 0.364\nsamples = 576\nvalue = [138, 438]'),
 Text(0.2097902097902098, 0.8846153846153846, 'sexo_num <= 0.5\ngini = 0.144\nsamples = 308\nvalue = [24, 284]'),
 Text(0.33085664335664333, 0.9230769230769231, 'True  '),
 Text(0.1258741258741259, 0.8076923076923077, 'edad <= 42.5\ngini = 0.206\nsamples = 137\nvalue = [16, 121]'),
 Text(0.08391608391608392, 0.7307692307692307, 'edad <= 41.5\ngini = 0.287\nsamples = 69\nvalue = [12, 57]'),
 Text(0.06993006993006994, 0.6538461538461539, 'edad <= 27.5\ngini = 0.216\nsamples = 65\nvalue = [8, 57]'),
 Text(0.04195804195804196, 0.5769230769230769, 'edad <= 23.5\ngini = 0.053\nsamples = 37\nvalue = [1, 36]'),
 Text(0.027972027972027972, 0.5, 'edad <= 21.5\ngini = 0.105\nsamples = 18\nvalue = [1, 17]'),
 Text(0.013986013986013986, 0.4230769230769231, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.04195804195804196, 0.4230769230769231, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.055944055944055944, 0.5, 'gini = 0.0\nsamples = 19\nvalue = [0, 19]'),
 Text(0.0979020979020979, 0.5769230769230769, 'edad <= 29.0\ngini = 0.375\nsamples = 28\nvalue = [7, 21]'),
 Text(0.08391608391608392, 0.5, 'gini = 0.49\nsamples = 7\nvalue = [4, 3]'),
 Text(0.11188811188811189, 0.5, 'edad <= 34.0\ngini = 0.245\nsamples = 21\nvalue = [3, 18]'),
 Text(0.0979020979020979, 0.4230769230769231, 'gini = 0.0\nsamples = 8\nvalue = [0, 8]'),
 Text(0.1258741258741259, 0.4230769230769231, 'edad <= 36.0\ngini = 0.355\nsamples = 13\nvalue = [3, 10]'),
 Text(0.11188811188811189, 0.34615384615384615, 'gini = 0.444\nsamples = 9\nvalue = [3, 6]'),
 Text(0.13986013986013987, 0.34615384615384615, 'gini = 0.0\nsamples = 4\nvalue = [0, 4]'),
 Text(0.0979020979020979, 0.6538461538461539, 'gini = 0.0\nsamples = 4\nvalue = [4, 0]'),
 Text(0.16783216783216784, 0.7307692307692307, 'edad <= 53.5\ngini = 0.111\nsamples = 68\nvalue = [4, 64]'),
 Text(0.13986013986013987, 0.6538461538461539, 'edad <= 49.0\ngini = 0.042\nsamples = 47\nvalue = [1, 46]'),
 Text(0.1258741258741259, 0.5769230769230769, 'gini = 0.0\nsamples = 24\nvalue = [0, 24]'),
 Text(0.15384615384615385, 0.5769230769230769, 'edad <= 50.5\ngini = 0.083\nsamples = 23\nvalue = [1, 22]'),
 Text(0.13986013986013987, 0.5, 'gini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.16783216783216784, 0.5, 'gini = 0.0\nsamples = 15\nvalue = [0, 15]'),
 Text(0.1958041958041958, 0.6538461538461539, 'edad <= 54.5\ngini = 0.245\nsamples = 21\nvalue = [3, 18]'),
 Text(0.18181818181818182, 0.5769230769230769, 'gini = 0.5\nsamples = 4\nvalue = [2, 2]'),
 Text(0.2097902097902098, 0.5769230769230769, 'edad <= 58.5\ngini = 0.111\nsamples = 17\nvalue = [1, 16]'),
 Text(0.1958041958041958, 0.5, 'gini = 0.0\nsamples = 13\nvalue = [0, 13]'),
 Text(0.22377622377622378, 0.5, 'edad <= 60.5\ngini = 0.375\nsamples = 4\nvalue = [1, 3]'),
 Text(0.2097902097902098, 0.4230769230769231, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.23776223776223776, 0.4230769230769231, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.2937062937062937, 0.8076923076923077, 'edad <= 46.0\ngini = 0.089\nsamples = 171\nvalue = [8, 163]'),
 Text(0.26573426573426573, 0.7307692307692307, 'edad <= 32.5\ngini = 0.058\nsamples = 133\nvalue = [4, 129]'),
 Text(0.2517482517482518, 0.6538461538461539, 'edad <= 21.5\ngini = 0.081\nsamples = 95\nvalue = [4, 91]'),
 Text(0.23776223776223776, 0.5769230769230769, 'gini = 0.0\nsamples = 26\nvalue = [0, 26]'),
 Text(0.26573426573426573, 0.5769230769230769, 'edad <= 23.0\ngini = 0.109\nsamples = 69\nvalue = [4, 65]'),
 Text(0.2517482517482518, 0.5, 'gini = 0.5\nsamples = 2\nvalue = [1, 1]'),
 Text(0.27972027972027974, 0.5, 'edad <= 31.5\ngini = 0.086\nsamples = 67\nvalue = [3, 64]'),
 Text(0.26573426573426573, 0.4230769230769231, 'edad <= 25.5\ngini = 0.065\nsamples = 59\nvalue = [2, 57]'),
 Text(0.23776223776223776, 0.34615384615384615, 'edad <= 24.5\ngini = 0.198\nsamples = 9\nvalue = [1, 8]'),
 Text(0.22377622377622378, 0.2692307692307692, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.2517482517482518, 0.2692307692307692, 'gini = 0.278\nsamples = 6\nvalue = [1, 5]'),
 Text(0.2937062937062937, 0.34615384615384615, 'edad <= 28.5\ngini = 0.039\nsamples = 50\nvalue = [1, 49]'),
 Text(0.27972027972027974, 0.2692307692307692, 'edad <= 27.5\ngini = 0.071\nsamples = 27\nvalue = [1, 26]'),
 Text(0.26573426573426573, 0.19230769230769232, 'gini = 0.0\nsamples = 7\nvalue = [0, 7]'),
 Text(0.2937062937062937, 0.19230769230769232, 'gini = 0.095\nsamples = 20\nvalue = [1, 19]'),
 Text(0.3076923076923077, 0.2692307692307692, 'gini = 0.0\nsamples = 23\nvalue = [0, 23]'),
 Text(0.2937062937062937, 0.4230769230769231, 'gini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.27972027972027974, 0.6538461538461539, 'gini = 0.0\nsamples = 38\nvalue = [0, 38]'),
 Text(0.32167832167832167, 0.7307692307692307, 'edad <= 48.0\ngini = 0.188\nsamples = 38\nvalue = [4, 34]'),
 Text(0.3076923076923077, 0.6538461538461539, 'gini = 0.494\nsamples = 9\nvalue = [4, 5]'),
 Text(0.3356643356643357, 0.6538461538461539, 'gini = 0.0\nsamples = 29\nvalue = [0, 29]'),
 Text(0.6940559440559441, 0.8846153846153846, 'edad <= 40.5\ngini = 0.489\nsamples = 268\nvalue = [114, 154]'),
 Text(0.5729895104895105, 0.9230769230769231, '  False'),
 Text(0.4991258741258741, 0.8076923076923077, 'sexo_num <= 0.5\ngini = 0.426\nsamples = 143\nvalue = [44, 99]'),
 Text(0.42482517482517484, 0.7307692307692307, 'edad <= 21.5\ngini = 0.461\nsamples = 114\nvalue = [41, 73]'),
 Text(0.36363636363636365, 0.6538461538461539, 'edad <= 19.5\ngini = 0.298\nsamples = 22\nvalue = [4, 18]'),
 Text(0.3356643356643357, 0.5769230769230769, 'edad <= 18.5\ngini = 0.444\nsamples = 3\nvalue = [2, 1]'),
 Text(0.32167832167832167, 0.5, 'gini = 0.0\nsamples = 1\nvalue = [0, 1]'),
 Text(0.34965034965034963, 0.5, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.3916083916083916, 0.5769230769230769, 'edad <= 20.5\ngini = 0.188\nsamples = 19\nvalue = [2, 17]'),
 Text(0.3776223776223776, 0.5, 'gini = 0.245\nsamples = 7\nvalue = [1, 6]'),
 Text(0.40559440559440557, 0.5, 'gini = 0.153\nsamples = 12\nvalue = [1, 11]'),
 Text(0.486013986013986, 0.6538461538461539, 'edad <= 23.5\ngini = 0.481\nsamples = 92\nvalue = [37, 55]'),
 Text(0.44755244755244755, 0.5769230769230769, 'edad <= 22.5\ngini = 0.48\nsamples = 15\nvalue = [9, 6]'),
 Text(0.43356643356643354, 0.5, 'gini = 0.48\nsamples = 10\nvalue = [6, 4]'),
 Text(0.46153846153846156, 0.5, 'gini = 0.48\nsamples = 5\nvalue = [3, 2]'),
 Text(0.5244755244755245, 0.5769230769230769, 'edad <= 25.5\ngini = 0.463\nsamples = 77\nvalue = [28, 49]'),
 Text(0.48951048951048953, 0.5, 'edad <= 24.5\ngini = 0.32\nsamples = 20\nvalue = [4, 16]'),
 Text(0.4755244755244755, 0.4230769230769231, 'gini = 0.42\nsamples = 10\nvalue = [3, 7]'),
 Text(0.5034965034965035, 0.4230769230769231, 'gini = 0.18\nsamples = 10\nvalue = [1, 9]'),
 Text(0.5594405594405595, 0.5, 'edad <= 29.5\ngini = 0.488\nsamples = 57\nvalue = [24, 33]'),
 Text(0.5314685314685315, 0.4230769230769231, 'edad <= 28.5\ngini = 0.219\nsamples = 8\nvalue = [7, 1]'),
 Text(0.5174825174825175, 0.34615384615384615, 'gini = 0.0\nsamples = 5\nvalue = [5, 0]'),
 Text(0.5454545454545454, 0.34615384615384615, 'gini = 0.444\nsamples = 3\nvalue = [2, 1]'),
 Text(0.5874125874125874, 0.4230769230769231, 'edad <= 31.5\ngini = 0.453\nsamples = 49\nvalue = [17.0, 32.0]'),
 Text(0.5734265734265734, 0.34615384615384615, 'gini = 0.0\nsamples = 2\nvalue = [0, 2]'),
 Text(0.6013986013986014, 0.34615384615384615, 'edad <= 36.5\ngini = 0.462\nsamples = 47\nvalue = [17, 30]'),
 Text(0.5664335664335665, 0.2692307692307692, 'edad <= 34.5\ngini = 0.478\nsamples = 33\nvalue = [13, 20]'),
 Text(0.5384615384615384, 0.19230769230769232, 'edad <= 33.5\ngini = 0.444\nsamples = 18\nvalue = [6, 12]'),
 Text(0.5244755244755245, 0.11538461538461539, 'edad <= 32.5\ngini = 0.457\nsamples = 17\nvalue = [6, 11]'),
 Text(0.5104895104895105, 0.038461538461538464, 'gini = 0.5\nsamples = 2\nvalue = [1, 1]'),
 Text(0.5384615384615384, 0.038461538461538464, 'gini = 0.444\nsamples = 15\nvalue = [5, 10]'),
 Text(0.5524475524475524, 0.11538461538461539, 'gini = 0.0\nsamples = 1\nvalue = [0, 1]'),
 Text(0.5944055944055944, 0.19230769230769232, 'edad <= 35.5\ngini = 0.498\nsamples = 15\nvalue = [7, 8]'),
 Text(0.5804195804195804, 0.11538461538461539, 'gini = 0.444\nsamples = 3\nvalue = [2, 1]'),
 Text(0.6083916083916084, 0.11538461538461539, 'gini = 0.486\nsamples = 12\nvalue = [5, 7]'),
 Text(0.6363636363636364, 0.2692307692307692, 'edad <= 38.5\ngini = 0.408\nsamples = 14\nvalue = [4, 10]'),
 Text(0.6223776223776224, 0.19230769230769232, 'gini = 0.346\nsamples = 9\nvalue = [2, 7]'),
 Text(0.6503496503496503, 0.19230769230769232, 'gini = 0.48\nsamples = 5\nvalue = [2, 3]'),
 Text(0.5734265734265734, 0.7307692307692307, 'edad <= 19.0\ngini = 0.185\nsamples = 29\nvalue = [3, 26]'),
 Text(0.5594405594405595, 0.6538461538461539, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.5874125874125874, 0.6538461538461539, 'edad <= 32.0\ngini = 0.071\nsamples = 27\nvalue = [1, 26]'),
 Text(0.5734265734265734, 0.5769230769230769, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.6013986013986014, 0.5769230769230769, 'edad <= 34.0\ngini = 0.18\nsamples = 10\nvalue = [1, 9]'),
 Text(0.5874125874125874, 0.5, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.6153846153846154, 0.5, 'gini = 0.0\nsamples = 9\nvalue = [0, 9]'),
 Text(0.888986013986014, 0.8076923076923077, 'edad <= 56.5\ngini = 0.493\nsamples = 125\nvalue = [70, 55]'),
 Text(0.833916083916084, 0.7307692307692307, 'edad <= 52.5\ngini = 0.462\nsamples = 91\nvalue = [58.0, 33.0]'),
 Text(0.7937062937062938, 0.6538461538461539, 'sexo_num <= 0.5\ngini = 0.5\nsamples = 47\nvalue = [23.0, 24.0]'),
 Text(0.7797202797202797, 0.5769230769230769, 'edad <= 48.5\ngini = 0.465\nsamples = 38\nvalue = [14, 24]'),
 Text(0.7132867132867133, 0.5, 'edad <= 43.5\ngini = 0.494\nsamples = 27\nvalue = [12, 15]'),
 Text(0.6643356643356644, 0.4230769230769231, 'edad <= 41.5\ngini = 0.444\nsamples = 9\nvalue = [3, 6]'),
 Text(0.6503496503496503, 0.34615384615384615, 'gini = 0.444\nsamples = 3\nvalue = [2, 1]'),
 Text(0.6783216783216783, 0.34615384615384615, 'edad <= 42.5\ngini = 0.278\nsamples = 6\nvalue = [1, 5]'),
 Text(0.6643356643356644, 0.2692307692307692, 'gini = 0.444\nsamples = 3\nvalue = [1, 2]'),
 Text(0.6923076923076923, 0.2692307692307692, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.7622377622377622, 0.4230769230769231, 'edad <= 45.5\ngini = 0.5\nsamples = 18\nvalue = [9, 9]'),
 Text(0.7342657342657343, 0.34615384615384615, 'edad <= 44.5\ngini = 0.444\nsamples = 6\nvalue = [4, 2]'),
 Text(0.7202797202797203, 0.2692307692307692, 'gini = 0.5\nsamples = 4\nvalue = [2, 2]'),
 Text(0.7482517482517482, 0.2692307692307692, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.7902097902097902, 0.34615384615384615, 'edad <= 47.0\ngini = 0.486\nsamples = 12\nvalue = [5, 7]'),
 Text(0.7762237762237763, 0.2692307692307692, 'gini = 0.32\nsamples = 5\nvalue = [1, 4]'),
 Text(0.8041958041958042, 0.2692307692307692, 'gini = 0.49\nsamples = 7\nvalue = [4, 3]'),
 Text(0.8461538461538461, 0.5, 'edad <= 51.0\ngini = 0.298\nsamples = 11\nvalue = [2, 9]'),
 Text(0.8321678321678322, 0.4230769230769231, 'edad <= 49.5\ngini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.8181818181818182, 0.34615384615384615, 'gini = 0.278\nsamples = 6\nvalue = [1, 5]'),
 Text(0.8461538461538461, 0.34615384615384615, 'gini = 0.0\nsamples = 2\nvalue = [0, 2]'),
 Text(0.8601398601398601, 0.4230769230769231, 'gini = 0.444\nsamples = 3\nvalue = [1, 2]'),
 Text(0.8076923076923077, 0.5769230769230769, 'gini = 0.0\nsamples = 9\nvalue = [9, 0]'),
 Text(0.8741258741258742, 0.6538461538461539, 'edad <= 53.5\ngini = 0.325\nsamples = 44\nvalue = [35, 9]'),
 Text(0.8601398601398601, 0.5769230769230769, 'gini = 0.0\nsamples = 6\nvalue = [6, 0]'),
 Text(0.8881118881118881, 0.5769230769230769, 'edad <= 54.5\ngini = 0.361\nsamples = 38\nvalue = [29, 9]'),
 Text(0.8741258741258742, 0.5, 'gini = 0.408\nsamples = 14\nvalue = [10, 4]'),
 Text(0.9020979020979021, 0.5, 'edad <= 55.5\ngini = 0.33\nsamples = 24\nvalue = [19, 5]'),
 Text(0.8881118881118881, 0.4230769230769231, 'gini = 0.198\nsamples = 9\nvalue = [8, 1]'),
 Text(0.916083916083916, 0.4230769230769231, 'gini = 0.391\nsamples = 15\nvalue = [11, 4]'),
 Text(0.9440559440559441, 0.7307692307692307, 'edad <= 58.0\ngini = 0.457\nsamples = 34\nvalue = [12, 22]'),
 Text(0.9300699300699301, 0.6538461538461539, 'gini = 0.278\nsamples = 18\nvalue = [3, 15]'),
 Text(0.958041958041958, 0.6538461538461539, 'edad <= 61.0\ngini = 0.492\nsamples = 16\nvalue = [9, 7]'),
 Text(0.9440559440559441, 0.5769230769230769, 'gini = 0.0\nsamples = 6\nvalue = [6, 0]'),
 Text(0.972027972027972, 0.5769230769230769, 'edad <= 64.5\ngini = 0.42\nsamples = 10\nvalue = [3, 7]'),
 Text(0.958041958041958, 0.5, 'edad <= 62.5\ngini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.9440559440559441, 0.4230769230769231, 'gini = 0.0\nsamples = 4\nvalue = [0, 4]'),
 Text(0.972027972027972, 0.4230769230769231, 'edad <= 63.5\ngini = 0.375\nsamples = 4\nvalue = [1, 3]'),
 Text(0.958041958041958, 0.34615384615384615, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.986013986013986, 0.34615384615384615, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.986013986013986, 0.5, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]')]

print(tree.tree_.max_depth)

12

print(tree.get_n_leaves())

69

from sklearn.metrics import accuracy_score, f1_score
yhat0 = tree.predict(X_test)
acc0 = accuracy_score(y_test, yhat0)
f10 = f1_score(y_test, yhat0, average='weighted')
print("Accuracy inicial: ", acc0)
print("F1-score inicial: ", f10)

Accuracy inicial:  0.8402777777777778
F1-score inicial:  0.8377859946121579

# Matriz de confusión - Árbol by Default.
from sklearn.metrics import confusion_matrix

conf_m = confusion_matrix(y_test, yhat0)
sns.heatmap(conf_m, annot = True, fmt = "d", cmap = "Blues", cbar = False, square = True)
plt.ylabel('y_true')
plt.xlabel('y_pred')
plt.title('Matriz de Confusión - Árbol Inicial.')
plt.show()

tree_smote = DTC().fit(X_train_smote, y_train_smote)

from sklearn.tree import plot_tree
plt.figure(figsize=(15,10))
plot_tree(tree_smote, filled = True, feature_names = X_train_smote.columns)

[Text(0.4680059523809524, 0.96875, 'institucion_unidad_medica_num <= 0.5\ngini = 0.5\nsamples = 876\nvalue = [438, 438]'),
 Text(0.2222222222222222, 0.90625, 'sexo_num <= 0.5\ngini = 0.38\nsamples = 381\nvalue = [97, 284]'),
 Text(0.3451140873015873, 0.9375, 'True  '),
 Text(0.09523809523809523, 0.84375, 'edad <= 21.5\ngini = 0.484\nsamples = 205\nvalue = [84, 121]'),
 Text(0.07936507936507936, 0.78125, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.1111111111111111, 0.78125, 'edad <= 49.5\ngini = 0.494\nsamples = 188\nvalue = [84, 104]'),
 Text(0.047619047619047616, 0.71875, 'edad <= 24.5\ngini = 0.498\nsamples = 136\nvalue = [72, 64]'),
 Text(0.031746031746031744, 0.65625, 'gini = 0.0\nsamples = 7\nvalue = [7, 0]'),
 Text(0.06349206349206349, 0.65625, 'edad <= 27.5\ngini = 0.5\nsamples = 129\nvalue = [65, 64]'),
 Text(0.031746031746031744, 0.59375, 'edad <= 25.5\ngini = 0.172\nsamples = 21\nvalue = [2, 19]'),
 Text(0.015873015873015872, 0.53125, 'gini = 0.408\nsamples = 7\nvalue = [2, 5]'),
 Text(0.047619047619047616, 0.53125, 'gini = 0.0\nsamples = 14\nvalue = [0, 14]'),
 Text(0.09523809523809523, 0.59375, 'edad <= 29.0\ngini = 0.486\nsamples = 108\nvalue = [63.0, 45.0]'),
 Text(0.07936507936507936, 0.53125, 'gini = 0.255\nsamples = 20\nvalue = [17, 3]'),
 Text(0.1111111111111111, 0.53125, 'edad <= 34.0\ngini = 0.499\nsamples = 88\nvalue = [46, 42]'),
 Text(0.07142857142857142, 0.46875, 'edad <= 32.5\ngini = 0.32\nsamples = 10\nvalue = [2, 8]'),
 Text(0.05555555555555555, 0.40625, 'edad <= 31.0\ngini = 0.5\nsamples = 4\nvalue = [2, 2]'),
 Text(0.03968253968253968, 0.34375, 'gini = 0.0\nsamples = 2\nvalue = [0, 2]'),
 Text(0.07142857142857142, 0.34375, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.0873015873015873, 0.40625, 'gini = 0.0\nsamples = 6\nvalue = [0, 6]'),
 Text(0.15079365079365079, 0.46875, 'edad <= 42.5\ngini = 0.492\nsamples = 78\nvalue = [44, 34]'),
 Text(0.11904761904761904, 0.40625, 'edad <= 41.5\ngini = 0.401\nsamples = 36\nvalue = [26, 10]'),
 Text(0.10317460317460317, 0.34375, 'edad <= 36.0\ngini = 0.473\nsamples = 26\nvalue = [16, 10]'),
 Text(0.0873015873015873, 0.28125, 'gini = 0.444\nsamples = 18\nvalue = [12, 6]'),
 Text(0.11904761904761904, 0.28125, 'edad <= 38.5\ngini = 0.5\nsamples = 8\nvalue = [4, 4]'),
 Text(0.10317460317460317, 0.21875, 'gini = 0.0\nsamples = 2\nvalue = [0, 2]'),
 Text(0.1349206349206349, 0.21875, 'edad <= 40.5\ngini = 0.444\nsamples = 6\nvalue = [4, 2]'),
 Text(0.11904761904761904, 0.15625, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.15079365079365079, 0.15625, 'gini = 0.5\nsamples = 4\nvalue = [2, 2]'),
 Text(0.1349206349206349, 0.34375, 'gini = 0.0\nsamples = 10\nvalue = [10, 0]'),
 Text(0.18253968253968253, 0.40625, 'edad <= 45.0\ngini = 0.49\nsamples = 42\nvalue = [18, 24]'),
 Text(0.16666666666666666, 0.34375, 'gini = 0.0\nsamples = 12\nvalue = [0, 12]'),
 Text(0.1984126984126984, 0.34375, 'edad <= 48.5\ngini = 0.48\nsamples = 30\nvalue = [18, 12]'),
 Text(0.18253968253968253, 0.28125, 'edad <= 46.5\ngini = 0.494\nsamples = 27\nvalue = [15, 12]'),
 Text(0.16666666666666666, 0.21875, 'gini = 0.375\nsamples = 4\nvalue = [3, 1]'),
 Text(0.1984126984126984, 0.21875, 'edad <= 47.5\ngini = 0.499\nsamples = 23\nvalue = [12, 11]'),
 Text(0.18253968253968253, 0.15625, 'gini = 0.498\nsamples = 17\nvalue = [8, 9]'),
 Text(0.21428571428571427, 0.15625, 'gini = 0.444\nsamples = 6\nvalue = [4, 2]'),
 Text(0.21428571428571427, 0.28125, 'gini = 0.0\nsamples = 3\nvalue = [3, 0]'),
 Text(0.1746031746031746, 0.71875, 'edad <= 53.5\ngini = 0.355\nsamples = 52\nvalue = [12, 40]'),
 Text(0.14285714285714285, 0.65625, 'edad <= 50.5\ngini = 0.083\nsamples = 23\nvalue = [1, 22]'),
 Text(0.12698412698412698, 0.59375, 'gini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.15873015873015872, 0.59375, 'gini = 0.0\nsamples = 15\nvalue = [0, 15]'),
 Text(0.20634920634920634, 0.65625, 'edad <= 54.5\ngini = 0.471\nsamples = 29\nvalue = [11, 18]'),
 Text(0.19047619047619047, 0.59375, 'gini = 0.298\nsamples = 11\nvalue = [9, 2]'),
 Text(0.2222222222222222, 0.59375, 'edad <= 58.5\ngini = 0.198\nsamples = 18\nvalue = [2, 16]'),
 Text(0.20634920634920634, 0.53125, 'gini = 0.0\nsamples = 13\nvalue = [0, 13]'),
 Text(0.23809523809523808, 0.53125, 'edad <= 60.5\ngini = 0.48\nsamples = 5\nvalue = [2, 3]'),
 Text(0.2222222222222222, 0.46875, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.25396825396825395, 0.46875, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.3492063492063492, 0.84375, 'edad <= 46.0\ngini = 0.137\nsamples = 176\nvalue = [13, 163]'),
 Text(0.31746031746031744, 0.78125, 'edad <= 32.5\ngini = 0.058\nsamples = 133\nvalue = [4, 129]'),
 Text(0.30158730158730157, 0.71875, 'edad <= 21.5\ngini = 0.081\nsamples = 95\nvalue = [4, 91]'),
 Text(0.2857142857142857, 0.65625, 'gini = 0.0\nsamples = 26\nvalue = [0, 26]'),
 Text(0.31746031746031744, 0.65625, 'edad <= 23.0\ngini = 0.109\nsamples = 69\nvalue = [4, 65]'),
 Text(0.30158730158730157, 0.59375, 'gini = 0.5\nsamples = 2\nvalue = [1, 1]'),
 Text(0.3333333333333333, 0.59375, 'edad <= 31.5\ngini = 0.086\nsamples = 67\nvalue = [3, 64]'),
 Text(0.31746031746031744, 0.53125, 'edad <= 25.5\ngini = 0.065\nsamples = 59\nvalue = [2, 57]'),
 Text(0.2857142857142857, 0.46875, 'edad <= 24.5\ngini = 0.198\nsamples = 9\nvalue = [1, 8]'),
 Text(0.2698412698412698, 0.40625, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.30158730158730157, 0.40625, 'gini = 0.278\nsamples = 6\nvalue = [1, 5]'),
 Text(0.3492063492063492, 0.46875, 'edad <= 28.5\ngini = 0.039\nsamples = 50\nvalue = [1, 49]'),
 Text(0.3333333333333333, 0.40625, 'edad <= 27.5\ngini = 0.071\nsamples = 27\nvalue = [1, 26]'),
 Text(0.31746031746031744, 0.34375, 'gini = 0.0\nsamples = 7\nvalue = [0, 7]'),
 Text(0.3492063492063492, 0.34375, 'gini = 0.095\nsamples = 20\nvalue = [1, 19]'),
 Text(0.36507936507936506, 0.40625, 'gini = 0.0\nsamples = 23\nvalue = [0, 23]'),
 Text(0.3492063492063492, 0.53125, 'gini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.3333333333333333, 0.71875, 'gini = 0.0\nsamples = 38\nvalue = [0, 38]'),
 Text(0.38095238095238093, 0.78125, 'edad <= 48.0\ngini = 0.331\nsamples = 43\nvalue = [9, 34]'),
 Text(0.36507936507936506, 0.71875, 'gini = 0.459\nsamples = 14\nvalue = [9, 5]'),
 Text(0.3968253968253968, 0.71875, 'gini = 0.0\nsamples = 29\nvalue = [0, 29]'),
 Text(0.7137896825396826, 0.90625, 'edad <= 42.5\ngini = 0.429\nsamples = 495\nvalue = [341, 154]'),
 Text(0.5908978174603174, 0.9375, '  False'),
 Text(0.5466269841269841, 0.84375, 'sexo_num <= 0.5\ngini = 0.493\nsamples = 232\nvalue = [130, 102]'),
 Text(0.4742063492063492, 0.78125, 'edad <= 19.5\ngini = 0.468\nsamples = 203\nvalue = [127, 76]'),
 Text(0.42857142857142855, 0.71875, 'edad <= 18.5\ngini = 0.124\nsamples = 15\nvalue = [14, 1]'),
 Text(0.4126984126984127, 0.65625, 'gini = 0.18\nsamples = 10\nvalue = [9, 1]'),
 Text(0.4444444444444444, 0.65625, 'gini = 0.0\nsamples = 5\nvalue = [5, 0]'),
 Text(0.5198412698412699, 0.71875, 'edad <= 21.5\ngini = 0.48\nsamples = 188\nvalue = [113, 75]'),
 Text(0.47619047619047616, 0.65625, 'edad <= 20.5\ngini = 0.188\nsamples = 19\nvalue = [2, 17]'),
 Text(0.4603174603174603, 0.59375, 'gini = 0.245\nsamples = 7\nvalue = [1, 6]'),
 Text(0.49206349206349204, 0.59375, 'gini = 0.153\nsamples = 12\nvalue = [1, 11]'),
 Text(0.5634920634920635, 0.65625, 'edad <= 23.5\ngini = 0.451\nsamples = 169\nvalue = [111.0, 58.0]'),
 Text(0.5238095238095238, 0.59375, 'edad <= 22.5\ngini = 0.26\nsamples = 39\nvalue = [33, 6]'),
 Text(0.5079365079365079, 0.53125, 'gini = 0.238\nsamples = 29\nvalue = [25, 4]'),
 Text(0.5396825396825397, 0.53125, 'gini = 0.32\nsamples = 10\nvalue = [8, 2]'),
 Text(0.6031746031746031, 0.59375, 'edad <= 25.5\ngini = 0.48\nsamples = 130\nvalue = [78, 52]'),
 Text(0.5714285714285714, 0.53125, 'edad <= 24.5\ngini = 0.473\nsamples = 26\nvalue = [10, 16]'),
 Text(0.5555555555555556, 0.46875, 'gini = 0.486\nsamples = 12\nvalue = [5, 7]'),
 Text(0.5873015873015873, 0.46875, 'gini = 0.459\nsamples = 14\nvalue = [5, 9]'),
 Text(0.6349206349206349, 0.53125, 'edad <= 28.5\ngini = 0.453\nsamples = 104\nvalue = [68, 36]'),
 Text(0.6190476190476191, 0.46875, 'gini = 0.0\nsamples = 12\nvalue = [12, 0]'),
 Text(0.6507936507936508, 0.46875, 'edad <= 41.5\ngini = 0.476\nsamples = 92\nvalue = [56, 36]'),
 Text(0.6349206349206349, 0.40625, 'edad <= 40.5\ngini = 0.472\nsamples = 89\nvalue = [55.0, 34.0]'),
 Text(0.6190476190476191, 0.34375, 'edad <= 36.5\ngini = 0.479\nsamples = 83\nvalue = [50.0, 33.0]'),
 Text(0.5793650793650794, 0.28125, 'edad <= 31.5\ngini = 0.46\nsamples = 64\nvalue = [41, 23]'),
 Text(0.5476190476190477, 0.21875, 'edad <= 29.5\ngini = 0.48\nsamples = 5\nvalue = [2, 3]'),
 Text(0.5317460317460317, 0.15625, 'gini = 0.444\nsamples = 3\nvalue = [2, 1]'),
 Text(0.5634920634920635, 0.15625, 'gini = 0.0\nsamples = 2\nvalue = [0, 2]'),
 Text(0.6111111111111112, 0.21875, 'edad <= 32.5\ngini = 0.448\nsamples = 59\nvalue = [39, 20]'),
 Text(0.5952380952380952, 0.15625, 'gini = 0.32\nsamples = 5\nvalue = [4, 1]'),
 Text(0.626984126984127, 0.15625, 'edad <= 34.5\ngini = 0.456\nsamples = 54\nvalue = [35, 19]'),
 Text(0.5952380952380952, 0.09375, 'edad <= 33.5\ngini = 0.471\nsamples = 29\nvalue = [18, 11]'),
 Text(0.5793650793650794, 0.03125, 'gini = 0.459\nsamples = 28\nvalue = [18, 10]'),
 Text(0.6111111111111112, 0.03125, 'gini = 0.0\nsamples = 1\nvalue = [0, 1]'),
 Text(0.6587301587301587, 0.09375, 'edad <= 35.5\ngini = 0.435\nsamples = 25\nvalue = [17, 8]'),
 Text(0.6428571428571429, 0.03125, 'gini = 0.278\nsamples = 6\nvalue = [5, 1]'),
 Text(0.6746031746031746, 0.03125, 'gini = 0.465\nsamples = 19\nvalue = [12, 7]'),
 Text(0.6587301587301587, 0.28125, 'edad <= 38.5\ngini = 0.499\nsamples = 19\nvalue = [9, 10]'),
 Text(0.6428571428571429, 0.21875, 'gini = 0.463\nsamples = 11\nvalue = [4, 7]'),
 Text(0.6746031746031746, 0.21875, 'gini = 0.469\nsamples = 8\nvalue = [5, 3]'),
 Text(0.6507936507936508, 0.34375, 'gini = 0.278\nsamples = 6\nvalue = [5, 1]'),
 Text(0.6666666666666666, 0.40625, 'gini = 0.444\nsamples = 3\nvalue = [1, 2]'),
 Text(0.6190476190476191, 0.78125, 'edad <= 19.0\ngini = 0.185\nsamples = 29\nvalue = [3, 26]'),
 Text(0.6031746031746031, 0.71875, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.6349206349206349, 0.71875, 'edad <= 32.0\ngini = 0.071\nsamples = 27\nvalue = [1, 26]'),
 Text(0.6190476190476191, 0.65625, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.6507936507936508, 0.65625, 'edad <= 34.0\ngini = 0.18\nsamples = 10\nvalue = [1, 9]'),
 Text(0.6349206349206349, 0.59375, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.6666666666666666, 0.59375, 'gini = 0.0\nsamples = 9\nvalue = [0, 9]'),
 Text(0.8809523809523809, 0.84375, 'edad <= 56.5\ngini = 0.317\nsamples = 263\nvalue = [211, 52]'),
 Text(0.8253968253968254, 0.78125, 'edad <= 52.5\ngini = 0.249\nsamples = 206\nvalue = [176.0, 30.0]'),
 Text(0.7936507936507936, 0.71875, 'sexo_num <= 0.5\ngini = 0.363\nsamples = 88\nvalue = [67, 21]'),
 Text(0.7777777777777778, 0.65625, 'edad <= 48.5\ngini = 0.458\nsamples = 59\nvalue = [38, 21]'),
 Text(0.746031746031746, 0.59375, 'edad <= 47.0\ngini = 0.397\nsamples = 44\nvalue = [32, 12]'),
 Text(0.7301587301587301, 0.53125, 'edad <= 45.5\ngini = 0.495\nsamples = 20\nvalue = [11, 9]'),
 Text(0.7142857142857143, 0.46875, 'edad <= 43.5\ngini = 0.444\nsamples = 15\nvalue = [10, 5]'),
 Text(0.6984126984126984, 0.40625, 'gini = 0.375\nsamples = 4\nvalue = [1, 3]'),
 Text(0.7301587301587301, 0.40625, 'edad <= 44.5\ngini = 0.298\nsamples = 11\nvalue = [9, 2]'),
 Text(0.7142857142857143, 0.34375, 'gini = 0.444\nsamples = 6\nvalue = [4, 2]'),
 Text(0.746031746031746, 0.34375, 'gini = 0.0\nsamples = 5\nvalue = [5, 0]'),
 Text(0.746031746031746, 0.46875, 'gini = 0.32\nsamples = 5\nvalue = [1, 4]'),
 Text(0.7619047619047619, 0.53125, 'gini = 0.219\nsamples = 24\nvalue = [21, 3]'),
 Text(0.8095238095238095, 0.59375, 'edad <= 51.0\ngini = 0.48\nsamples = 15\nvalue = [6, 9]'),
 Text(0.7936507936507936, 0.53125, 'edad <= 49.5\ngini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.7777777777777778, 0.46875, 'gini = 0.278\nsamples = 6\nvalue = [1, 5]'),
 Text(0.8095238095238095, 0.46875, 'gini = 0.0\nsamples = 2\nvalue = [0, 2]'),
 Text(0.8253968253968254, 0.53125, 'gini = 0.408\nsamples = 7\nvalue = [5, 2]'),
 Text(0.8095238095238095, 0.65625, 'gini = 0.0\nsamples = 29\nvalue = [29, 0]'),
 Text(0.8571428571428571, 0.71875, 'edad <= 53.5\ngini = 0.141\nsamples = 118\nvalue = [109, 9]'),
 Text(0.8412698412698413, 0.65625, 'gini = 0.0\nsamples = 22\nvalue = [22, 0]'),
 Text(0.873015873015873, 0.65625, 'edad <= 54.5\ngini = 0.17\nsamples = 96\nvalue = [87, 9]'),
 Text(0.8571428571428571, 0.59375, 'gini = 0.208\nsamples = 34\nvalue = [30, 4]'),
 Text(0.8888888888888888, 0.59375, 'edad <= 55.5\ngini = 0.148\nsamples = 62\nvalue = [57, 5]'),
 Text(0.873015873015873, 0.53125, 'gini = 0.08\nsamples = 24\nvalue = [23, 1]'),
 Text(0.9047619047619048, 0.53125, 'gini = 0.188\nsamples = 38\nvalue = [34, 4]'),
 Text(0.9365079365079365, 0.78125, 'edad <= 58.0\ngini = 0.474\nsamples = 57\nvalue = [35, 22]'),
 Text(0.9206349206349206, 0.71875, 'gini = 0.332\nsamples = 19\nvalue = [4, 15]'),
 Text(0.9523809523809523, 0.71875, 'edad <= 61.5\ngini = 0.301\nsamples = 38\nvalue = [31, 7]'),
 Text(0.9365079365079365, 0.65625, 'gini = 0.0\nsamples = 23\nvalue = [23, 0]'),
 Text(0.9682539682539683, 0.65625, 'edad <= 64.5\ngini = 0.498\nsamples = 15\nvalue = [8, 7]'),
 Text(0.9523809523809523, 0.59375, 'edad <= 63.5\ngini = 0.497\nsamples = 13\nvalue = [6, 7]'),
 Text(0.9365079365079365, 0.53125, 'edad <= 62.5\ngini = 0.494\nsamples = 9\nvalue = [5, 4]'),
 Text(0.9206349206349206, 0.46875, 'gini = 0.49\nsamples = 7\nvalue = [3, 4]'),
 Text(0.9523809523809523, 0.46875, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.9682539682539683, 0.53125, 'gini = 0.375\nsamples = 4\nvalue = [1, 3]'),
 Text(0.9841269841269841, 0.59375, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]')]

print(tree_smote.tree_.max_depth)

15

print(tree_smote.get_n_leaves())

78

from sklearn.metrics import accuracy_score, f1_score
yhat0 = tree_smote.predict(X_test)
acc0 = accuracy_score(y_test, yhat0)
f10 = f1_score(y_test, yhat0, average='weighted')
print("Accuracy inicial: ", acc0)
print("F1-score inicial: ", f10)

Accuracy inicial:  0.8125
F1-score inicial:  0.8179125376992675

# Matriz de confusión - Árbol inicial.
from sklearn.metrics import confusion_matrix

conf_m = confusion_matrix(y_test, yhat0)
sns.heatmap(conf_m, annot = True, fmt = "d", cmap = "Blues", cbar = False, square = True)
plt.ylabel('y_true')
plt.xlabel('y_pred')
plt.title('Matriz de Confusión - Árbol Inicial.')
plt.show()

from sklearn.model_selection import cross_val_score, StratifiedKFold
skf = StratifiedKFold(n_splits = 4)
ccp = np.linspace(0.001, 0.2, 250)
cv_scores = []
for alpha in ccp:
    pruned_tree = DTC(ccp_alpha = alpha, class_weight='balanced')
    cv_scores.append(np.mean(cross_val_score(pruned_tree, X_train, y_train, cv = skf, scoring = 'f1')))

alpha = ccp[np.argmax(cv_scores)]
print("Best alpha: ", alpha)

Best alpha:  0.001

pruned_tree = DTC(ccp_alpha = alpha).fit(X_train, y_train)
plot_tree(pruned_tree, filled = True, feature_names = X_train.columns)

[Text(0.4642857142857143, 0.95, 'institucion_unidad_medica_num <= 0.5\ngini = 0.364\nsamples = 576\nvalue = [138, 438]'),
 Text(0.25, 0.85, 'sexo_num <= 0.5\ngini = 0.144\nsamples = 308\nvalue = [24, 284]'),
 Text(0.35714285714285715, 0.8999999999999999, 'True  '),
 Text(0.19047619047619047, 0.75, 'edad <= 42.5\ngini = 0.206\nsamples = 137\nvalue = [16, 121]'),
 Text(0.14285714285714285, 0.65, 'edad <= 41.5\ngini = 0.287\nsamples = 69\nvalue = [12, 57]'),
 Text(0.11904761904761904, 0.55, 'edad <= 27.5\ngini = 0.216\nsamples = 65\nvalue = [8, 57]'),
 Text(0.07142857142857142, 0.45, 'edad <= 23.5\ngini = 0.053\nsamples = 37\nvalue = [1, 36]'),
 Text(0.047619047619047616, 0.35, 'edad <= 21.5\ngini = 0.105\nsamples = 18\nvalue = [1, 17]'),
 Text(0.023809523809523808, 0.25, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.07142857142857142, 0.25, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.09523809523809523, 0.35, 'gini = 0.0\nsamples = 19\nvalue = [0, 19]'),
 Text(0.16666666666666666, 0.45, 'edad <= 29.0\ngini = 0.375\nsamples = 28\nvalue = [7, 21]'),
 Text(0.14285714285714285, 0.35, 'gini = 0.49\nsamples = 7\nvalue = [4, 3]'),
 Text(0.19047619047619047, 0.35, 'gini = 0.245\nsamples = 21\nvalue = [3, 18]'),
 Text(0.16666666666666666, 0.55, 'gini = 0.0\nsamples = 4\nvalue = [4, 0]'),
 Text(0.23809523809523808, 0.65, 'edad <= 53.5\ngini = 0.111\nsamples = 68\nvalue = [4, 64]'),
 Text(0.21428571428571427, 0.55, 'gini = 0.042\nsamples = 47\nvalue = [1, 46]'),
 Text(0.2619047619047619, 0.55, 'edad <= 54.5\ngini = 0.245\nsamples = 21\nvalue = [3, 18]'),
 Text(0.23809523809523808, 0.45, 'gini = 0.5\nsamples = 4\nvalue = [2, 2]'),
 Text(0.2857142857142857, 0.45, 'edad <= 58.5\ngini = 0.111\nsamples = 17\nvalue = [1, 16]'),
 Text(0.2619047619047619, 0.35, 'gini = 0.0\nsamples = 13\nvalue = [0, 13]'),
 Text(0.30952380952380953, 0.35, 'edad <= 60.5\ngini = 0.375\nsamples = 4\nvalue = [1, 3]'),
 Text(0.2857142857142857, 0.25, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.3333333333333333, 0.25, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.30952380952380953, 0.75, 'edad <= 46.0\ngini = 0.089\nsamples = 171\nvalue = [8, 163]'),
 Text(0.2857142857142857, 0.65, 'gini = 0.058\nsamples = 133\nvalue = [4, 129]'),
 Text(0.3333333333333333, 0.65, 'edad <= 48.0\ngini = 0.188\nsamples = 38\nvalue = [4, 34]'),
 Text(0.30952380952380953, 0.55, 'gini = 0.494\nsamples = 9\nvalue = [4, 5]'),
 Text(0.35714285714285715, 0.55, 'gini = 0.0\nsamples = 29\nvalue = [0, 29]'),
 Text(0.6785714285714286, 0.85, 'edad <= 40.5\ngini = 0.489\nsamples = 268\nvalue = [114, 154]'),
 Text(0.5714285714285714, 0.8999999999999999, '  False'),
 Text(0.5119047619047619, 0.75, 'sexo_num <= 0.5\ngini = 0.426\nsamples = 143\nvalue = [44, 99]'),
 Text(0.4523809523809524, 0.65, 'edad <= 21.5\ngini = 0.461\nsamples = 114\nvalue = [41, 73]'),
 Text(0.40476190476190477, 0.55, 'edad <= 19.5\ngini = 0.298\nsamples = 22\nvalue = [4, 18]'),
 Text(0.38095238095238093, 0.45, 'edad <= 18.5\ngini = 0.444\nsamples = 3\nvalue = [2, 1]'),
 Text(0.35714285714285715, 0.35, 'gini = 0.0\nsamples = 1\nvalue = [0, 1]'),
 Text(0.40476190476190477, 0.35, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.42857142857142855, 0.45, 'gini = 0.188\nsamples = 19\nvalue = [2, 17]'),
 Text(0.5, 0.55, 'edad <= 23.5\ngini = 0.481\nsamples = 92\nvalue = [37, 55]'),
 Text(0.47619047619047616, 0.45, 'gini = 0.48\nsamples = 15\nvalue = [9, 6]'),
 Text(0.5238095238095238, 0.45, 'edad <= 25.5\ngini = 0.463\nsamples = 77\nvalue = [28, 49]'),
 Text(0.5, 0.35, 'gini = 0.32\nsamples = 20\nvalue = [4, 16]'),
 Text(0.5476190476190477, 0.35, 'edad <= 29.5\ngini = 0.488\nsamples = 57\nvalue = [24, 33]'),
 Text(0.5238095238095238, 0.25, 'gini = 0.219\nsamples = 8\nvalue = [7, 1]'),
 Text(0.5714285714285714, 0.25, 'gini = 0.453\nsamples = 49\nvalue = [17.0, 32.0]'),
 Text(0.5714285714285714, 0.65, 'edad <= 19.0\ngini = 0.185\nsamples = 29\nvalue = [3, 26]'),
 Text(0.5476190476190477, 0.55, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.5952380952380952, 0.55, 'edad <= 32.0\ngini = 0.071\nsamples = 27\nvalue = [1, 26]'),
 Text(0.5714285714285714, 0.45, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.6190476190476191, 0.45, 'edad <= 34.0\ngini = 0.18\nsamples = 10\nvalue = [1, 9]'),
 Text(0.5952380952380952, 0.35, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.6428571428571429, 0.35, 'gini = 0.0\nsamples = 9\nvalue = [0, 9]'),
 Text(0.8452380952380952, 0.75, 'edad <= 56.5\ngini = 0.493\nsamples = 125\nvalue = [70, 55]'),
 Text(0.7857142857142857, 0.65, 'edad <= 52.5\ngini = 0.462\nsamples = 91\nvalue = [58.0, 33.0]'),
 Text(0.7380952380952381, 0.55, 'sexo_num <= 0.5\ngini = 0.5\nsamples = 47\nvalue = [23.0, 24.0]'),
 Text(0.7142857142857143, 0.45, 'edad <= 48.5\ngini = 0.465\nsamples = 38\nvalue = [14, 24]'),
 Text(0.6904761904761905, 0.35, 'edad <= 43.5\ngini = 0.494\nsamples = 27\nvalue = [12, 15]'),
 Text(0.6309523809523809, 0.25, 'edad <= 41.5\ngini = 0.444\nsamples = 9\nvalue = [3, 6]'),
 Text(0.6071428571428571, 0.15, 'gini = 0.444\nsamples = 3\nvalue = [2, 1]'),
 Text(0.6547619047619048, 0.15, 'gini = 0.278\nsamples = 6\nvalue = [1, 5]'),
 Text(0.75, 0.25, 'edad <= 45.5\ngini = 0.5\nsamples = 18\nvalue = [9, 9]'),
 Text(0.7023809523809523, 0.15, 'edad <= 44.5\ngini = 0.444\nsamples = 6\nvalue = [4, 2]'),
 Text(0.6785714285714286, 0.05, 'gini = 0.5\nsamples = 4\nvalue = [2, 2]'),
 Text(0.7261904761904762, 0.05, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.7976190476190477, 0.15, 'edad <= 47.0\ngini = 0.486\nsamples = 12\nvalue = [5, 7]'),
 Text(0.7738095238095238, 0.05, 'gini = 0.32\nsamples = 5\nvalue = [1, 4]'),
 Text(0.8214285714285714, 0.05, 'gini = 0.49\nsamples = 7\nvalue = [4, 3]'),
 Text(0.7380952380952381, 0.35, 'gini = 0.298\nsamples = 11\nvalue = [2, 9]'),
 Text(0.7619047619047619, 0.45, 'gini = 0.0\nsamples = 9\nvalue = [9, 0]'),
 Text(0.8333333333333334, 0.55, 'edad <= 53.5\ngini = 0.325\nsamples = 44\nvalue = [35, 9]'),
 Text(0.8095238095238095, 0.45, 'gini = 0.0\nsamples = 6\nvalue = [6, 0]'),
 Text(0.8571428571428571, 0.45, 'gini = 0.361\nsamples = 38\nvalue = [29, 9]'),
 Text(0.9047619047619048, 0.65, 'edad <= 58.0\ngini = 0.457\nsamples = 34\nvalue = [12, 22]'),
 Text(0.8809523809523809, 0.55, 'gini = 0.278\nsamples = 18\nvalue = [3, 15]'),
 Text(0.9285714285714286, 0.55, 'edad <= 61.0\ngini = 0.492\nsamples = 16\nvalue = [9, 7]'),
 Text(0.9047619047619048, 0.45, 'gini = 0.0\nsamples = 6\nvalue = [6, 0]'),
 Text(0.9523809523809523, 0.45, 'edad <= 64.5\ngini = 0.42\nsamples = 10\nvalue = [3, 7]'),
 Text(0.9285714285714286, 0.35, 'edad <= 62.5\ngini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.9047619047619048, 0.25, 'gini = 0.0\nsamples = 4\nvalue = [0, 4]'),
 Text(0.9523809523809523, 0.25, 'edad <= 63.5\ngini = 0.375\nsamples = 4\nvalue = [1, 3]'),
 Text(0.9285714285714286, 0.15, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.9761904761904762, 0.15, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.9761904761904762, 0.35, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]')]

print(pruned_tree.tree_.max_depth)

9

print(pruned_tree.get_n_leaves())

41

yhat_p = pruned_tree.predict(X_test)
acc_p = accuracy_score(y_test, yhat_p)
f1_p = f1_score(y_test, yhat_p, average = 'weighted')
print("Accuracy final: ", acc_p)
print("F1-score final: ", f1_p)

Accuracy final:  0.8611111111111112
F1-score final:  0.8546087480572208

# Matriz de confusión - Árbol inicial
from sklearn.metrics import confusion_matrix

conf_m = confusion_matrix(y_test, yhat_p)
sns.heatmap(conf_m, annot=True, fmt="d", cmap="Blues", cbar=False, square=True)
plt.ylabel('y_true')
plt.xlabel('y_pred')
plt.title('Matriz de Confusión - Árbol Podado.')
plt.show()

from sklearn.model_selection import cross_val_score, StratifiedKFold
skf = StratifiedKFold(n_splits = 4)
ccp = np.linspace(0.001, 0.2, 250)
cv_scores = []
for alpha in ccp:
    pruned_tree = DTC(ccp_alpha = alpha, class_weight='balanced')
    cv_scores.append(np.mean(cross_val_score(pruned_tree, X_train_smote, y_train_smote, cv = skf, scoring = 'f1')))

alpha = ccp[np.argmax(cv_scores)]
print("Best alpha: ", alpha)

Best alpha:  0.001

pruned_tree = DTC(ccp_alpha = alpha).fit(X_train_smote, y_train_smote)
plot_tree(pruned_tree, filled = True, feature_names = X_train_smote.columns)

[Text(0.4253472222222222, 0.9583333333333334, 'institucion_unidad_medica_num <= 0.5\ngini = 0.5\nsamples = 876\nvalue = [438, 438]'),
 Text(0.19791666666666666, 0.875, 'sexo_num <= 0.5\ngini = 0.38\nsamples = 381\nvalue = [97, 284]'),
 Text(0.3116319444444444, 0.9166666666666667, 'True  '),
 Text(0.11805555555555555, 0.7916666666666666, 'edad <= 21.5\ngini = 0.484\nsamples = 205\nvalue = [84, 121]'),
 Text(0.09027777777777778, 0.7083333333333334, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.14583333333333334, 0.7083333333333334, 'edad <= 49.5\ngini = 0.494\nsamples = 188\nvalue = [84, 104]'),
 Text(0.06944444444444445, 0.625, 'edad <= 24.5\ngini = 0.498\nsamples = 136\nvalue = [72, 64]'),
 Text(0.041666666666666664, 0.5416666666666666, 'gini = 0.0\nsamples = 7\nvalue = [7, 0]'),
 Text(0.09722222222222222, 0.5416666666666666, 'edad <= 27.5\ngini = 0.5\nsamples = 129\nvalue = [65, 64]'),
 Text(0.06944444444444445, 0.4583333333333333, 'gini = 0.172\nsamples = 21\nvalue = [2, 19]'),
 Text(0.125, 0.4583333333333333, 'edad <= 29.0\ngini = 0.486\nsamples = 108\nvalue = [63.0, 45.0]'),
 Text(0.09722222222222222, 0.375, 'gini = 0.255\nsamples = 20\nvalue = [17, 3]'),
 Text(0.1527777777777778, 0.375, 'edad <= 34.0\ngini = 0.499\nsamples = 88\nvalue = [46, 42]'),
 Text(0.08333333333333333, 0.2916666666666667, 'edad <= 32.5\ngini = 0.32\nsamples = 10\nvalue = [2, 8]'),
 Text(0.05555555555555555, 0.20833333333333334, 'edad <= 31.0\ngini = 0.5\nsamples = 4\nvalue = [2, 2]'),
 Text(0.027777777777777776, 0.125, 'gini = 0.0\nsamples = 2\nvalue = [0, 2]'),
 Text(0.08333333333333333, 0.125, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.1111111111111111, 0.20833333333333334, 'gini = 0.0\nsamples = 6\nvalue = [0, 6]'),
 Text(0.2222222222222222, 0.2916666666666667, 'edad <= 42.5\ngini = 0.492\nsamples = 78\nvalue = [44, 34]'),
 Text(0.16666666666666666, 0.20833333333333334, 'edad <= 41.5\ngini = 0.401\nsamples = 36\nvalue = [26, 10]'),
 Text(0.1388888888888889, 0.125, 'gini = 0.473\nsamples = 26\nvalue = [16, 10]'),
 Text(0.19444444444444445, 0.125, 'gini = 0.0\nsamples = 10\nvalue = [10, 0]'),
 Text(0.2777777777777778, 0.20833333333333334, 'edad <= 45.0\ngini = 0.49\nsamples = 42\nvalue = [18, 24]'),
 Text(0.25, 0.125, 'gini = 0.0\nsamples = 12\nvalue = [0, 12]'),
 Text(0.3055555555555556, 0.125, 'edad <= 48.5\ngini = 0.48\nsamples = 30\nvalue = [18, 12]'),
 Text(0.2777777777777778, 0.041666666666666664, 'gini = 0.494\nsamples = 27\nvalue = [15, 12]'),
 Text(0.3333333333333333, 0.041666666666666664, 'gini = 0.0\nsamples = 3\nvalue = [3, 0]'),
 Text(0.2222222222222222, 0.625, 'edad <= 53.5\ngini = 0.355\nsamples = 52\nvalue = [12, 40]'),
 Text(0.19444444444444445, 0.5416666666666666, 'gini = 0.083\nsamples = 23\nvalue = [1, 22]'),
 Text(0.25, 0.5416666666666666, 'edad <= 54.5\ngini = 0.471\nsamples = 29\nvalue = [11, 18]'),
 Text(0.2222222222222222, 0.4583333333333333, 'gini = 0.298\nsamples = 11\nvalue = [9, 2]'),
 Text(0.2777777777777778, 0.4583333333333333, 'edad <= 58.5\ngini = 0.198\nsamples = 18\nvalue = [2, 16]'),
 Text(0.25, 0.375, 'gini = 0.0\nsamples = 13\nvalue = [0, 13]'),
 Text(0.3055555555555556, 0.375, 'edad <= 60.5\ngini = 0.48\nsamples = 5\nvalue = [2, 3]'),
 Text(0.2777777777777778, 0.2916666666666667, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.3333333333333333, 0.2916666666666667, 'gini = 0.0\nsamples = 3\nvalue = [0, 3]'),
 Text(0.2777777777777778, 0.7916666666666666, 'edad <= 46.0\ngini = 0.137\nsamples = 176\nvalue = [13, 163]'),
 Text(0.25, 0.7083333333333334, 'gini = 0.058\nsamples = 133\nvalue = [4, 129]'),
 Text(0.3055555555555556, 0.7083333333333334, 'edad <= 48.0\ngini = 0.331\nsamples = 43\nvalue = [9, 34]'),
 Text(0.2777777777777778, 0.625, 'gini = 0.459\nsamples = 14\nvalue = [9, 5]'),
 Text(0.3333333333333333, 0.625, 'gini = 0.0\nsamples = 29\nvalue = [0, 29]'),
 Text(0.6527777777777778, 0.875, 'edad <= 42.5\ngini = 0.429\nsamples = 495\nvalue = [341, 154]'),
 Text(0.5390625, 0.9166666666666667, '  False'),
 Text(0.4722222222222222, 0.7916666666666666, 'sexo_num <= 0.5\ngini = 0.493\nsamples = 232\nvalue = [130, 102]'),
 Text(0.4166666666666667, 0.7083333333333334, 'edad <= 19.5\ngini = 0.468\nsamples = 203\nvalue = [127, 76]'),
 Text(0.3888888888888889, 0.625, 'gini = 0.124\nsamples = 15\nvalue = [14, 1]'),
 Text(0.4444444444444444, 0.625, 'edad <= 21.5\ngini = 0.48\nsamples = 188\nvalue = [113, 75]'),
 Text(0.4166666666666667, 0.5416666666666666, 'gini = 0.188\nsamples = 19\nvalue = [2, 17]'),
 Text(0.4722222222222222, 0.5416666666666666, 'edad <= 23.5\ngini = 0.451\nsamples = 169\nvalue = [111.0, 58.0]'),
 Text(0.4444444444444444, 0.4583333333333333, 'gini = 0.26\nsamples = 39\nvalue = [33, 6]'),
 Text(0.5, 0.4583333333333333, 'edad <= 25.5\ngini = 0.48\nsamples = 130\nvalue = [78, 52]'),
 Text(0.4722222222222222, 0.375, 'gini = 0.473\nsamples = 26\nvalue = [10, 16]'),
 Text(0.5277777777777778, 0.375, 'edad <= 28.5\ngini = 0.453\nsamples = 104\nvalue = [68, 36]'),
 Text(0.5, 0.2916666666666667, 'gini = 0.0\nsamples = 12\nvalue = [12, 0]'),
 Text(0.5555555555555556, 0.2916666666666667, 'gini = 0.476\nsamples = 92\nvalue = [56, 36]'),
 Text(0.5277777777777778, 0.7083333333333334, 'edad <= 19.0\ngini = 0.185\nsamples = 29\nvalue = [3, 26]'),
 Text(0.5, 0.625, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]'),
 Text(0.5555555555555556, 0.625, 'edad <= 32.0\ngini = 0.071\nsamples = 27\nvalue = [1, 26]'),
 Text(0.5277777777777778, 0.5416666666666666, 'gini = 0.0\nsamples = 17\nvalue = [0, 17]'),
 Text(0.5833333333333334, 0.5416666666666666, 'edad <= 34.0\ngini = 0.18\nsamples = 10\nvalue = [1, 9]'),
 Text(0.5555555555555556, 0.4583333333333333, 'gini = 0.0\nsamples = 1\nvalue = [1, 0]'),
 Text(0.6111111111111112, 0.4583333333333333, 'gini = 0.0\nsamples = 9\nvalue = [0, 9]'),
 Text(0.8333333333333334, 0.7916666666666666, 'edad <= 56.5\ngini = 0.317\nsamples = 263\nvalue = [211, 52]'),
 Text(0.7777777777777778, 0.7083333333333334, 'edad <= 52.5\ngini = 0.249\nsamples = 206\nvalue = [176.0, 30.0]'),
 Text(0.75, 0.625, 'sexo_num <= 0.5\ngini = 0.363\nsamples = 88\nvalue = [67, 21]'),
 Text(0.7222222222222222, 0.5416666666666666, 'edad <= 48.5\ngini = 0.458\nsamples = 59\nvalue = [38, 21]'),
 Text(0.6666666666666666, 0.4583333333333333, 'edad <= 47.0\ngini = 0.397\nsamples = 44\nvalue = [32, 12]'),
 Text(0.6388888888888888, 0.375, 'edad <= 45.5\ngini = 0.495\nsamples = 20\nvalue = [11, 9]'),
 Text(0.6111111111111112, 0.2916666666666667, 'edad <= 43.5\ngini = 0.444\nsamples = 15\nvalue = [10, 5]'),
 Text(0.5833333333333334, 0.20833333333333334, 'gini = 0.375\nsamples = 4\nvalue = [1, 3]'),
 Text(0.6388888888888888, 0.20833333333333334, 'gini = 0.298\nsamples = 11\nvalue = [9, 2]'),
 Text(0.6666666666666666, 0.2916666666666667, 'gini = 0.32\nsamples = 5\nvalue = [1, 4]'),
 Text(0.6944444444444444, 0.375, 'gini = 0.219\nsamples = 24\nvalue = [21, 3]'),
 Text(0.7777777777777778, 0.4583333333333333, 'edad <= 51.0\ngini = 0.48\nsamples = 15\nvalue = [6, 9]'),
 Text(0.75, 0.375, 'gini = 0.219\nsamples = 8\nvalue = [1, 7]'),
 Text(0.8055555555555556, 0.375, 'gini = 0.408\nsamples = 7\nvalue = [5, 2]'),
 Text(0.7777777777777778, 0.5416666666666666, 'gini = 0.0\nsamples = 29\nvalue = [29, 0]'),
 Text(0.8055555555555556, 0.625, 'gini = 0.141\nsamples = 118\nvalue = [109, 9]'),
 Text(0.8888888888888888, 0.7083333333333334, 'edad <= 58.0\ngini = 0.474\nsamples = 57\nvalue = [35, 22]'),
 Text(0.8611111111111112, 0.625, 'gini = 0.332\nsamples = 19\nvalue = [4, 15]'),
 Text(0.9166666666666666, 0.625, 'edad <= 61.5\ngini = 0.301\nsamples = 38\nvalue = [31, 7]'),
 Text(0.8888888888888888, 0.5416666666666666, 'gini = 0.0\nsamples = 23\nvalue = [23, 0]'),
 Text(0.9444444444444444, 0.5416666666666666, 'edad <= 64.5\ngini = 0.498\nsamples = 15\nvalue = [8, 7]'),
 Text(0.9166666666666666, 0.4583333333333333, 'gini = 0.497\nsamples = 13\nvalue = [6, 7]'),
 Text(0.9722222222222222, 0.4583333333333333, 'gini = 0.0\nsamples = 2\nvalue = [2, 0]')]

print(pruned_tree.tree_.max_depth)

11

print(pruned_tree.get_n_leaves())

42

yhat_p = pruned_tree.predict(X_test)
acc_p = accuracy_score(y_test, yhat_p)
f1_p = f1_score(y_test, yhat_p, average = 'weighted')
print("Accuracy final: ", acc_p)
print("F1-score final: ", f1_p)

Accuracy final:  0.7916666666666666
F1-score final:  0.8018790849673203

# Matriz de confusión - Árbol inicial
from sklearn.metrics import confusion_matrix

conf_m = confusion_matrix(y_test, yhat_p)
sns.heatmap(conf_m, annot=True, fmt="d", cmap="Blues", cbar=False, square=True)
plt.ylabel('y_true')
plt.xlabel('y_pred')
plt.title('Matriz de Confusión - Árbol Podado.')
plt.show()

# Definimos entrada y salida.
X = df2[['edad', 'sexo_num', 'institucion_unidad_medica_num']]
y = df2['descripcion_grupo_enfermedad_num']
print(X.shape)
print(y.shape)

(720, 3)
(720,)

# Dividir los datos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0, stratify = y)
print(y_train.value_counts())
X_train.shape, X_test.shape, y_train.shape, y_test.shape

descripcion_grupo_enfermedad_num
1    438
0    138
Name: count, dtype: int64

((576, 3), (144, 3), (576,), (144,))

# Definir el modelo by Default.
model1 = RandomForestClassifier(random_state = 0)
# Entrenemos el modelo FS.
model1.fit(X_train, y_train)

RandomForestClassifier(random_state=0)

# Validar si el modelo pronostica adecuadamente.
y_pred_test = model1.predict(X_test)
print(y_pred_test[0:5])
print(y_test.head())

[1 0 0 1 1]
46382    0
32275    0
7409     1
32248    1
46407    1
Name: descripcion_grupo_enfermedad_num, dtype: int64

accuracy_train = model1.score(X_train, y_train)
print('Accuracy train = {:.2f}'.format(accuracy_train))
accuracy_test = model1.score(X_test, y_test)
print('Accuracy test = {:.2f}'.format(accuracy_test))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train-accuracy_test)*100))

Accuracy train = 0.88
Accuracy test = 0.86
Diferencia = 2.0833%

predRF = model1.predict(X_test)
repDT = classification_report(y_test, predRF)
print(repDT)

from imblearn.metrics import geometric_mean_score
print('G-mean =', geometric_mean_score(y_test, predRF))

disp = ConfusionMatrixDisplay.from_predictions(y_test, predRF, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.78      0.60      0.68        35
           1       0.88      0.94      0.91       109

    accuracy                           0.86       144
   macro avg       0.83      0.77      0.79       144
weighted avg       0.86      0.86      0.85       144

G-mean = 0.7529757479920719

from imblearn.over_sampling import SMOTE

# Definir la técnica.
smote = SMOTE(random_state = 0)
# Aplicamos.
X_smote, y_smote = smote.fit_resample(X, y)

# Comprobar si funicona.
print('Tamaño de X antes de SMOTE:', X.shape)
print('Tamaño de X después de SMOTE:', X_smote.shape)
print('Balance de clases con SMOTE:', y_smote.value_counts())
print('Nuestras clases están balanceadas.')

Tamaño de X antes de SMOTE: (720, 3)
Tamaño de X después de SMOTE: (1094, 3)
Balance de clases con SMOTE: descripcion_grupo_enfermedad_num
1    547
0    547
Name: count, dtype: int64
Nuestras clases están balanceadas.

# Dividir los tratos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X_smote, y_smote, test_size = 0.2, random_state = 0, stratify = y_smote)
print(y_train.value_counts())
X_train.shape, X_test.shape, y_train.shape, y_test.shape
# No dividir de manera que se desbalanceen los datos -> cuidado con el random_state).

descripcion_grupo_enfermedad_num
1    438
0    437
Name: count, dtype: int64

((875, 3), (219, 3), (875,), (219,))

# Definir el modelo balanceado.
model2 = RandomForestClassifier(random_state = 0)
# Entrenemos el modelo FS.
model2.fit(X_smote, y_smote)

RandomForestClassifier(random_state=0)

# Validar si el modelo pronostica adecuadamente.
y_pred_test = model2.predict(X_test)
print(y_pred_test[0:5])
print(y_test.head())

[0 1 0 0 0]
944     0
199     1
1034    0
405     0
727     0
Name: descripcion_grupo_enfermedad_num, dtype: int64

accuracy_train = model2.score(X_train, y_train)
print('Accuracy train = {:.2f}'.format(accuracy_train))
accuracy_test = model2.score(X_test, y_test)
print('Accuracy test = {:.2f}'.format(accuracy_test))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train-accuracy_test)*100))

Accuracy train = 0.87
Accuracy test = 0.90
Diferencia = 2.7543%

predRF = model2.predict(X_test)
repDT = classification_report(y_test, predRF)
print(repDT)

from imblearn.metrics import geometric_mean_score
print('G-mean =', geometric_mean_score(y_test, predRF))

disp = ConfusionMatrixDisplay.from_predictions(y_test, predRF, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.89      0.91      0.90       110
           1       0.91      0.89      0.90       109

    accuracy                           0.90       219
   macro avg       0.90      0.90      0.90       219
weighted avg       0.90      0.90      0.90       219

G-mean = 0.8994484455794076

from sklearn.svm import SVC

# Definimos entrada y salida.
X = df2[['edad', 'sexo_num', 'institucion_unidad_medica_num']]
y = df2['descripcion_grupo_enfermedad_num']

# Dividimos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0, stratify = y)
print(y_train.value_counts())
X_train.shape, X_test.shape

descripcion_grupo_enfermedad_num
1    438
0    138
Name: count, dtype: int64

((576, 3), (144, 3))

# Escalamos las características (recomendado para SVM).
scaler_svm = StandardScaler()
X_train_svm = scaler_svm.fit_transform(X_train)
X_test_svm = scaler_svm.transform(X_test)

# Definir el modelo by Default.
model_svm1 = SVC(kernel = 'rbf', random_state = 0)
# Entrenemos el modelo SVM.
model_svm1.fit(X_train_svm, y_train)

SVC(random_state=0)

accuracy_train_svm1 = model_svm1.score(X_train_svm, y_train)
print('Accuracy train = {:.2f}'.format(accuracy_train_svm1))
accuracy_test_svm1 = model_svm1.score(X_test_svm, y_test)
print('Accuracy test = {:.2f}'.format(accuracy_test_svm1))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train_svm1-accuracy_test_svm1)*100))

Accuracy train = 0.79
Accuracy test = 0.81
Diferencia = 1.9097%

pred_svm1 = model_svm1.predict(X_test_svm)
rep_svm1 = classification_report(y_test, pred_svm1)
print(rep_svm1)

from imblearn.metrics import geometric_mean_score
print('G-mean =', geometric_mean_score(y_test, pred_svm1))

disp = ConfusionMatrixDisplay.from_predictions(y_test, pred_svm1, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.64      0.51      0.57        35
           1       0.85      0.91      0.88       109

    accuracy                           0.81       144
   macro avg       0.75      0.71      0.73       144
weighted avg       0.80      0.81      0.81       144

G-mean = 0.6834497338233234

from imblearn.over_sampling import SMOTE

# Aplicamos SMOTE sobre los datos originales.
smote = SMOTE(random_state=0)
X_smote_svm, y_smote_svm = smote.fit_resample(X, y)

print('Tamaño de X antes de SMOTE:', X.shape)
print('Tamaño de X después de SMOTE:', X_smote_svm.shape)
print('Balance de clases con SMOTE:')
print(y_smote_svm.value_counts())
print('Nuestras clases están balanceadas.')

Tamaño de X antes de SMOTE: (720, 3)
Tamaño de X después de SMOTE: (1094, 3)
Balance de clases con SMOTE:
descripcion_grupo_enfermedad_num
1    547
0    547
Name: count, dtype: int64
Nuestras clases están balanceadas.

# Dividir los datos balanceados en train y test.
X_train, X_test, y_train, y_test = train_test_split(X_smote_svm, y_smote_svm, test_size=0.2, random_state=0, stratify=y_smote_svm)
print(y_train.value_counts())
X_train.shape, X_test.shape

descripcion_grupo_enfermedad_num
1    438
0    437
Name: count, dtype: int64

((875, 3), (219, 3))

# Escalamos las características.
scaler_svm2 = StandardScaler()
X_train_svm2 = scaler_svm2.fit_transform(X_train)
X_test_svm2 = scaler_svm2.transform(X_test)

# Definir el modelo balanceado.
model_svm2 = SVC(kernel = 'rbf', random_state = 0)
# Entrenemos el modelo SVM.
model_svm2.fit(X_train_svm2, y_train)

SVC(random_state=0)

# Evaluación del modelo SVM balanceado.
accuracy_train_svm2 = model_svm2.score(X_train_svm2, y_train)
print('Accuracy train = {:.2f}'.format(accuracy_train_svm2))
accuracy_test_svm2 = model_svm2.score(X_test_svm2, y_test)
print('Accuracy test = {:.2f}'.format(accuracy_test_svm2))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train_svm2-accuracy_test_svm2)*100))

Accuracy train = 0.75
Accuracy test = 0.75
Diferencia = 0.8282%

pred_svm2 = model_svm2.predict(X_test_svm2)
rep_svm2 = classification_report(y_test, pred_svm2)
print(rep_svm2)

print('G-mean =', geometric_mean_score(y_test, pred_svm2))

disp = ConfusionMatrixDisplay.from_predictions(y_test, pred_svm2, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.73      0.82      0.77       110
           1       0.79      0.69      0.74       109

    accuracy                           0.75       219
   macro avg       0.76      0.75      0.75       219
weighted avg       0.76      0.75      0.75       219

G-mean = 0.7503126954482326

# Activación y perdida.
def relu(z):
    return np.maximum(0, z)

def relu_deriv(z):
    return (z > 0).astype(float)

def sigmoid(z):
    return 1/(1+np.exp(-np.clip(z, -500, 500)))

def binary_cross_entropy(y_true, y_pred):
    eps = 1e-8
    return -np.mean(y_true*np.log(y_pred+eps)+(1-y_true)*np.log(1 - y_pred+eps))


# Generación de la Red Neuronal.
class RedNeuronal:
    """Red neuronal binaria de una capa oculta construida desde cero con NumPy."""

    def __init__(self, n_entrada = 3, n_oculta = 6, lr = 0.01, epochs = 1000, random_state = 0):
        self.n_entrada = n_entrada
        self.n_oculta = n_oculta
        self.lr = lr
        self.epochs = epochs
        self.random_state = random_state
        self.losses = []

    def _init_params(self):
        rng = np.random.default_rng(self.random_state)
        # He initialization (recomendada para ReLU).
        self.W1 = rng.standard_normal((self.n_entrada, self.n_oculta))*np.sqrt(2/self.n_entrada)
        self.b1 = np.zeros((1, self.n_oculta))
        self.W2 = rng.standard_normal((self.n_oculta, 1))*np.sqrt(2/self.n_oculta)
        self.b2 = np.zeros((1, 1))

    def _forward(self, X):
        self.Z1 = X @ self.W1+self.b1 # (n, n_oculta).
        self.A1 = relu(self.Z1) # (n, n_oculta).
        self.Z2 = self.A1 @ self.W2+self.b2 # (n, 1).
        self.A2 = sigmoid(self.Z2) # (n, 1).
        return self.A2

    def _backward(self, X, y):
        n = X.shape[0]
        # Gradiente capa de salida.
        dZ2 = self.A2-y # (n, 1).
        dW2 = (self.A1.T @ dZ2)/n
        db2 = np.mean(dZ2, axis = 0, keepdims = True)
        # Gradiente capa oculta.
        dA1 = dZ2 @ self.W2.T # (n, n_oculta).
        dZ1 = dA1*relu_deriv(self.Z1)
        dW1 = (X.T @ dZ1)/n
        db1 = np.mean(dZ1, axis = 0, keepdims = True)
        # Actualizar pesos.
        self.W2 -= self.lr*dW2
        self.b2 -= self.lr*db2
        self.W1 -= self.lr*dW1
        self.b1 -= self.lr*db1

    def fit(self, X, y):
        self._init_params()
        y = y.reshape(-1, 1).astype(float)
        for epoch in range(self.epochs):
            y_pred = self._forward(X)
            loss = binary_cross_entropy(y, y_pred)
            self.losses.append(loss)
            self._backward(X, y)
            if (epoch+1)%200 == 0:
                print(f'Epoch {epoch+1}/{self.epochs}  |  Loss: {loss:.4f}')
        return self

    def predict_proba(self, X):
        return self._forward(X).flatten()

    def predict(self, X, threshold=0.5):
        return (self.predict_proba(X) >= threshold).astype(int)

    def score(self, X, y):
        return accuracy_score(y, self.predict(X))

    def plot_loss(self, title='Curva de pérdida'):
        plt.figure(figsize=(7, 3))
        plt.plot(self.losses)
        plt.xlabel('Época')
        plt.ylabel('Binary Cross-Entropy')
        plt.title(title)
        plt.tight_layout()
        plt.show()

print('Clase RedNeuronal definida correctamente.')

Clase RedNeuronal definida correctamente.

# Definimos entrada y salida.
X = df2[['edad', 'sexo_num', 'institucion_unidad_medica_num']]
y = df2['descripcion_grupo_enfermedad_num']

# Dividimos en train y test.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0, stratify=y)
print(y_train.value_counts())
X_train.shape, X_test.shape

descripcion_grupo_enfermedad_num
1    438
0    138
Name: count, dtype: int64

((576, 3), (144, 3))

# Escalamos las características (necesario para la red neuronal).
scaler_nn1 = StandardScaler()
X_train_nn1 = scaler_nn1.fit_transform(X_train)
X_test_nn1 = scaler_nn1.transform(X_test)

# Definir el modelo by Default.
model_nn1 = RedNeuronal(n_entrada = 3, n_oculta = 6, lr = 0.01, epochs = 1000, random_state = 0)
# Entrenamos el modelo.
model_nn1.fit(X_train_nn1, y_train.values)

Epoch 200/1000  |  Loss: 0.4644
Epoch 400/1000  |  Loss: 0.4563
Epoch 600/1000  |  Loss: 0.4531
Epoch 800/1000  |  Loss: 0.4518
Epoch 1000/1000  |  Loss: 0.4511

<__main__.RedNeuronal at 0x31f27d6a0>

# Curva de pérdida.
model_nn1.plot_loss(title='Curva de pérdida - Red Neuronal Desbalanceada')

# Evaluación del modelo.
accuracy_train_nn1 = model_nn1.score(X_train_nn1, y_train.values)
print('Accuracy train = {:.2f}'.format(accuracy_train_nn1))
accuracy_test_nn1 = model_nn1.score(X_test_nn1, y_test.values)
print('Accuracy test = {:.2f}'.format(accuracy_test_nn1))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train_nn1-accuracy_test_nn1)*100))

Accuracy train = 0.76
Accuracy test = 0.76
Diferencia = 0.3472%

pred_nn1 = model_nn1.predict(X_test_nn1)
rep_nn1  = classification_report(y_test, pred_nn1)
print(rep_nn1)

print('G-mean =', geometric_mean_score(y_test, pred_nn1))

disp = ConfusionMatrixDisplay.from_predictions(y_test, pred_nn1, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.00      0.00      0.00        35
           1       0.76      1.00      0.86       109

    accuracy                           0.76       144
   macro avg       0.38      0.50      0.43       144
weighted avg       0.57      0.76      0.65       144

G-mean = 0.0

/opt/anaconda3/lib/python3.13/site-packages/sklearn/metrics/_classification.py:1833: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, f"{metric.capitalize()} is", result.shape[0])
/opt/anaconda3/lib/python3.13/site-packages/sklearn/metrics/_classification.py:1833: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, f"{metric.capitalize()} is", result.shape[0])
/opt/anaconda3/lib/python3.13/site-packages/sklearn/metrics/_classification.py:1833: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.
  _warn_prf(average, modifier, f"{metric.capitalize()} is", result.shape[0])

from imblearn.over_sampling import SMOTE

# Aplicamos SMOTE sobre los datos originales.
smote = SMOTE(random_state=0)
X_smote_nn, y_smote_nn = smote.fit_resample(X, y)

print('Tamaño de X antes de SMOTE:', X.shape)
print('Tamaño de X después de SMOTE:', X_smote_nn.shape)
print('Balance de clases con SMOTE:')
print(y_smote_nn.value_counts())
print('Nuestras clases están balanceadas.')

Tamaño de X antes de SMOTE: (720, 3)
Tamaño de X después de SMOTE: (1094, 3)
Balance de clases con SMOTE:
descripcion_grupo_enfermedad_num
1    547
0    547
Name: count, dtype: int64
Nuestras clases están balanceadas.

# Dividir los datos balanceados en train y test.
X_train, X_test, y_train, y_test = train_test_split(X_smote_nn, y_smote_nn, test_size = 0.2, random_state = 0, stratify = y_smote_nn)
print(y_train.value_counts())
X_train.shape, X_test.shape

descripcion_grupo_enfermedad_num
1    438
0    437
Name: count, dtype: int64

((875, 3), (219, 3))

# Escalamos las características.
scaler_nn2 = StandardScaler()
X_train_nn2 = scaler_nn2.fit_transform(X_train)
X_test_nn2 = scaler_nn2.transform(X_test)

# Definir el modelo balanceado.
model_nn2 = RedNeuronal(n_entrada = 3, n_oculta = 6, lr = 0.01, epochs = 1000, random_state = 0)
# Entrenemos el modelo.
model_nn2.fit(X_train_nn2, y_train.values)

Epoch 200/1000  |  Loss: 0.5910
Epoch 400/1000  |  Loss: 0.5781
Epoch 600/1000  |  Loss: 0.5706
Epoch 800/1000  |  Loss: 0.5658
Epoch 1000/1000  |  Loss: 0.5624

<__main__.RedNeuronal at 0x327c92850>

# Curva de pérdida.
model_nn2.plot_loss(title='Curva de pérdida - Red Neuronal Balanceada (SMOTE)')

# Evaluación del modelo.
accuracy_train_nn2 = model_nn2.score(X_train_nn2, y_train.values)
print('Accuracy train = {:.2f}'.format(accuracy_train_nn2))
accuracy_test_nn2 = model_nn2.score(X_test_nn2, y_test.values)
print('Accuracy test = {:.2f}'.format(accuracy_test_nn2))
print('Diferencia = {:.4f}%'.format(np.abs(accuracy_train_nn2-accuracy_test_nn2)*100))

Accuracy train = 0.72
Accuracy test = 0.74
Diferencia = 2.6578%

pred_nn2 = model_nn2.predict(X_test_nn2)
rep_nn2  = classification_report(y_test, pred_nn2)
print(rep_nn2)

print('G-mean =', geometric_mean_score(y_test, pred_nn2))

disp = ConfusionMatrixDisplay.from_predictions(y_test, pred_nn2, cmap = plt.cm.Blues)

# 0: Factores influyendo en el estado de salud. 1: Trastorno mental y del comportamiento.

              precision    recall  f1-score   support

           0       0.72      0.80      0.76       110
           1       0.77      0.69      0.73       109

    accuracy                           0.74       219
   macro avg       0.75      0.74      0.74       219
weighted avg       0.75      0.74      0.74       219

G-mean = 0.7419290502442469

Dep. Variable:	descripcion_grupo_enfermedad_num	R-squared:	0.557
Model:	OLS	Adj. R-squared:	0.553
Method:	Least Squares	F-statistic:	149.4
Date:	Tue, 24 Mar 2026	Prob (F-statistic):	1.74e-122
Time:	23:52:19	Log-Likelihood:	-116.25
No. Observations:	720	AIC:	246.5
Df Residuals:	713	BIC:	278.6
Df Model:	6
Covariance Type:	nonrobust

	coef	std err	t	P>\|t\|	[0.025	0.975]
Intercept	1.2893	0.078	16.470	0.000	1.136	1.443
edad	-0.0005	0.001	-0.575	0.565	-0.002	0.001
sexo_num	-0.1472	0.026	-5.559	0.000	-0.199	-0.095
IMC	-0.0018	0.002	-0.876	0.382	-0.006	0.002
institucion_unidad_medica_num	-0.0496	0.026	-1.882	0.060	-0.101	0.002
clave_enfermedad_num	-0.0189	0.001	-17.611	0.000	-0.021	-0.017
descripcion_enfermedad_num	0.0056	0.001	5.564	0.000	0.004	0.008

Omnibus:	758.401	Durbin-Watson:	1.802
Prob(Omnibus):	0.000	Jarque-Bera (JB):	46.874
Skew:	0.010	Prob(JB):	6.63e-11
Kurtosis:	1.750	Cond. No.	473.

	penalty penalty: {'l1', 'l2', 'elasticnet', None}, default='l2' Specify the norm of the penalty: - `None`: no penalty is added; - `'l2'`: add a L2 penalty term and it is the default choice; - `'l1'`: add a L1 penalty term; - `'elasticnet'`: both L1 and L2 penalty terms are added. .. warning:: Some penalties may not work with some solvers. See the parameter `solver` below, to know the compatibility between the penalty and solver. .. versionadded:: 0.19 l1 penalty with SAGA solver (allowing 'multinomial' + L1) .. deprecated:: 1.8 `penalty` was deprecated in version 1.8 and will be removed in 1.10. Use `l1_ratio` instead. `l1_ratio=0` for `penalty='l2'`, `l1_ratio=1` for `penalty='l1'` and `l1_ratio` set to any float between 0 and 1 for `'penalty='elasticnet'`.	'deprecated'
	C C: float, default=1.0 Inverse of regularization strength; must be a positive float. Like in support vector machines, smaller values specify stronger regularization. `C=np.inf` results in unpenalized logistic regression. For a visual example on the effect of tuning the `C` parameter with an L1 penalty, see: :ref:`sphx_glr_auto_examples_linear_model_plot_logistic_path.py`.	1.0
	l1_ratio l1_ratio: float, default=0.0 The Elastic-Net mixing parameter, with `0 <= l1_ratio <= 1`. Setting `l1_ratio=1` gives a pure L1-penalty, setting `l1_ratio=0` a pure L2-penalty. Any value between 0 and 1 gives an Elastic-Net penalty of the form `l1_ratio * L1 + (1 - l1_ratio) * L2`. .. warning:: Certain values of `l1_ratio`, i.e. some penalties, may not work with some solvers. See the parameter `solver` below, to know the compatibility between the penalty and solver. .. versionchanged:: 1.8 Default value changed from None to 0.0. .. deprecated:: 1.8 `None` is deprecated and will be removed in version 1.10. Always use `l1_ratio` to specify the penalty type.	0.0
	dual dual: bool, default=False Dual (constrained) or primal (regularized, see also :ref:`this equation `) formulation. Dual formulation is only implemented for l2 penalty with liblinear solver. Prefer `dual=False` when n_samples > n_features.	False
	tol tol: float, default=1e-4 Tolerance for stopping criteria.	0.0001
	fit_intercept fit_intercept: bool, default=True Specifies if a constant (a.k.a. bias or intercept) should be added to the decision function.	True
	intercept_scaling intercept_scaling: float, default=1 Useful only when the solver `liblinear` is used and `self.fit_intercept` is set to `True`. In this case, `x` becomes `[x, self.intercept_scaling]`, i.e. a "synthetic" feature with constant value equal to `intercept_scaling` is appended to the instance vector. The intercept becomes ``intercept_scaling * synthetic_feature_weight``. .. note:: The synthetic feature weight is subject to L1 or L2 regularization as all other features. To lessen the effect of regularization on synthetic feature weight (and therefore on the intercept) `intercept_scaling` has to be increased.	1
	class_weight class_weight: dict or 'balanced', default=None Weights associated with classes in the form ``{class_label: weight}``. If not given, all classes are supposed to have weight one. The "balanced" mode uses the values of y to automatically adjust weights inversely proportional to class frequencies in the input data as ``n_samples / (n_classes * np.bincount(y))``. Note that these weights will be multiplied with sample_weight (passed through the fit method) if sample_weight is specified. .. versionadded:: 0.17 class_weight='balanced'	None
	random_state random_state: int, RandomState instance, default=None Used when ``solver`` == 'sag', 'saga' or 'liblinear' to shuffle the data. See :term:`Glossary ` for details.	0
	solver solver: {'lbfgs', 'liblinear', 'newton-cg', 'newton-cholesky', 'sag', 'saga'}, default='lbfgs' Algorithm to use in the optimization problem. Default is 'lbfgs'. To choose a solver, you might want to consider the following aspects: - 'lbfgs' is a good default solver because it works reasonably well for a wide class of problems. - For :term:`multiclass` problems (`n_classes >= 3`), all solvers except 'liblinear' minimize the full multinomial loss, 'liblinear' will raise an error. - 'newton-cholesky' is a good choice for `n_samples` >> `n_features * n_classes`, especially with one-hot encoded categorical features with rare categories. Be aware that the memory usage of this solver has a quadratic dependency on `n_features * n_classes` because it explicitly computes the full Hessian matrix. - For small datasets, 'liblinear' is a good choice, whereas 'sag' and 'saga' are faster for large ones; - 'liblinear' can only handle binary classification by default. To apply a one-versus-rest scheme for the multiclass setting one can wrap it with the :class:`~sklearn.multiclass.OneVsRestClassifier`. .. warning:: The choice of the algorithm depends on the penalty chosen (`l1_ratio=0` for L2-penalty, `l1_ratio=1` for L1-penalty and `0 < l1_ratio < 1` for Elastic-Net) and on (multinomial) multiclass support: ================= ======================== ====================== solver l1_ratio multinomial multiclass ================= ======================== ====================== 'lbfgs' l1_ratio=0 yes 'liblinear' l1_ratio=1 or l1_ratio=0 no 'newton-cg' l1_ratio=0 yes 'newton-cholesky' l1_ratio=0 yes 'sag' l1_ratio=0 yes 'saga' 0<=l1_ratio<=1 yes ================= ======================== ====================== .. note:: 'sag' and 'saga' fast convergence is only guaranteed on features with approximately the same scale. You can preprocess the data with a scaler from :mod:`sklearn.preprocessing`. .. seealso:: Refer to the :ref:`User Guide ` for more information regarding :class:`LogisticRegression` and more specifically the :ref:`Table ` summarizing solver/penalty supports. .. versionadded:: 0.17 Stochastic Average Gradient (SAG) descent solver. Multinomial support in version 0.18. .. versionadded:: 0.19 SAGA solver. .. versionchanged:: 0.22 The default solver changed from 'liblinear' to 'lbfgs' in 0.22. .. versionadded:: 1.2 newton-cholesky solver. Multinomial support in version 1.6.	'lbfgs'
	max_iter max_iter: int, default=100 Maximum number of iterations taken for the solvers to converge.	100
	verbose verbose: int, default=0 For the liblinear and lbfgs solvers set verbose to any positive number for verbosity.	0
	warm_start warm_start: bool, default=False When set to True, reuse the solution of the previous call to fit as initialization, otherwise, just erase the previous solution. Useless for liblinear solver. See :term:`the Glossary `. .. versionadded:: 0.17 warm_start to support lbfgs, newton-cg, sag, saga solvers.	False
	n_jobs n_jobs: int, default=None Does not have any effect. .. deprecated:: 1.8 `n_jobs` is deprecated in version 1.8 and will be removed in 1.10.	None

	penalty penalty: {'l1', 'l2', 'elasticnet', None}, default='l2' Specify the norm of the penalty: - `None`: no penalty is added; - `'l2'`: add a L2 penalty term and it is the default choice; - `'l1'`: add a L1 penalty term; - `'elasticnet'`: both L1 and L2 penalty terms are added. .. warning:: Some penalties may not work with some solvers. See the parameter `solver` below, to know the compatibility between the penalty and solver. .. versionadded:: 0.19 l1 penalty with SAGA solver (allowing 'multinomial' + L1) .. deprecated:: 1.8 `penalty` was deprecated in version 1.8 and will be removed in 1.10. Use `l1_ratio` instead. `l1_ratio=0` for `penalty='l2'`, `l1_ratio=1` for `penalty='l1'` and `l1_ratio` set to any float between 0 and 1 for `'penalty='elasticnet'`.	'deprecated'
	C C: float, default=1.0 Inverse of regularization strength; must be a positive float. Like in support vector machines, smaller values specify stronger regularization. `C=np.inf` results in unpenalized logistic regression. For a visual example on the effect of tuning the `C` parameter with an L1 penalty, see: :ref:`sphx_glr_auto_examples_linear_model_plot_logistic_path.py`.	1.0
	l1_ratio l1_ratio: float, default=0.0 The Elastic-Net mixing parameter, with `0 <= l1_ratio <= 1`. Setting `l1_ratio=1` gives a pure L1-penalty, setting `l1_ratio=0` a pure L2-penalty. Any value between 0 and 1 gives an Elastic-Net penalty of the form `l1_ratio * L1 + (1 - l1_ratio) * L2`. .. warning:: Certain values of `l1_ratio`, i.e. some penalties, may not work with some solvers. See the parameter `solver` below, to know the compatibility between the penalty and solver. .. versionchanged:: 1.8 Default value changed from None to 0.0. .. deprecated:: 1.8 `None` is deprecated and will be removed in version 1.10. Always use `l1_ratio` to specify the penalty type.	0.0
	dual dual: bool, default=False Dual (constrained) or primal (regularized, see also :ref:`this equation `) formulation. Dual formulation is only implemented for l2 penalty with liblinear solver. Prefer `dual=False` when n_samples > n_features.	False
	tol tol: float, default=1e-4 Tolerance for stopping criteria.	0.0001
	fit_intercept fit_intercept: bool, default=True Specifies if a constant (a.k.a. bias or intercept) should be added to the decision function.	True
	intercept_scaling intercept_scaling: float, default=1 Useful only when the solver `liblinear` is used and `self.fit_intercept` is set to `True`. In this case, `x` becomes `[x, self.intercept_scaling]`, i.e. a "synthetic" feature with constant value equal to `intercept_scaling` is appended to the instance vector. The intercept becomes ``intercept_scaling * synthetic_feature_weight``. .. note:: The synthetic feature weight is subject to L1 or L2 regularization as all other features. To lessen the effect of regularization on synthetic feature weight (and therefore on the intercept) `intercept_scaling` has to be increased.	1
	class_weight class_weight: dict or 'balanced', default=None Weights associated with classes in the form ``{class_label: weight}``. If not given, all classes are supposed to have weight one. The "balanced" mode uses the values of y to automatically adjust weights inversely proportional to class frequencies in the input data as ``n_samples / (n_classes * np.bincount(y))``. Note that these weights will be multiplied with sample_weight (passed through the fit method) if sample_weight is specified. .. versionadded:: 0.17 class_weight='balanced'	None
	random_state random_state: int, RandomState instance, default=None Used when ``solver`` == 'sag', 'saga' or 'liblinear' to shuffle the data. See :term:`Glossary ` for details.	0
	solver solver: {'lbfgs', 'liblinear', 'newton-cg', 'newton-cholesky', 'sag', 'saga'}, default='lbfgs' Algorithm to use in the optimization problem. Default is 'lbfgs'. To choose a solver, you might want to consider the following aspects: - 'lbfgs' is a good default solver because it works reasonably well for a wide class of problems. - For :term:`multiclass` problems (`n_classes >= 3`), all solvers except 'liblinear' minimize the full multinomial loss, 'liblinear' will raise an error. - 'newton-cholesky' is a good choice for `n_samples` >> `n_features * n_classes`, especially with one-hot encoded categorical features with rare categories. Be aware that the memory usage of this solver has a quadratic dependency on `n_features * n_classes` because it explicitly computes the full Hessian matrix. - For small datasets, 'liblinear' is a good choice, whereas 'sag' and 'saga' are faster for large ones; - 'liblinear' can only handle binary classification by default. To apply a one-versus-rest scheme for the multiclass setting one can wrap it with the :class:`~sklearn.multiclass.OneVsRestClassifier`. .. warning:: The choice of the algorithm depends on the penalty chosen (`l1_ratio=0` for L2-penalty, `l1_ratio=1` for L1-penalty and `0 < l1_ratio < 1` for Elastic-Net) and on (multinomial) multiclass support: ================= ======================== ====================== solver l1_ratio multinomial multiclass ================= ======================== ====================== 'lbfgs' l1_ratio=0 yes 'liblinear' l1_ratio=1 or l1_ratio=0 no 'newton-cg' l1_ratio=0 yes 'newton-cholesky' l1_ratio=0 yes 'sag' l1_ratio=0 yes 'saga' 0<=l1_ratio<=1 yes ================= ======================== ====================== .. note:: 'sag' and 'saga' fast convergence is only guaranteed on features with approximately the same scale. You can preprocess the data with a scaler from :mod:`sklearn.preprocessing`. .. seealso:: Refer to the :ref:`User Guide ` for more information regarding :class:`LogisticRegression` and more specifically the :ref:`Table ` summarizing solver/penalty supports. .. versionadded:: 0.17 Stochastic Average Gradient (SAG) descent solver. Multinomial support in version 0.18. .. versionadded:: 0.19 SAGA solver. .. versionchanged:: 0.22 The default solver changed from 'liblinear' to 'lbfgs' in 0.22. .. versionadded:: 1.2 newton-cholesky solver. Multinomial support in version 1.6.	'lbfgs'
	max_iter max_iter: int, default=100 Maximum number of iterations taken for the solvers to converge.	100
	verbose verbose: int, default=0 For the liblinear and lbfgs solvers set verbose to any positive number for verbosity.	0
	warm_start warm_start: bool, default=False When set to True, reuse the solution of the previous call to fit as initialization, otherwise, just erase the previous solution. Useless for liblinear solver. See :term:`the Glossary `. .. versionadded:: 0.17 warm_start to support lbfgs, newton-cg, sag, saga solvers.	False
	n_jobs n_jobs: int, default=None Does not have any effect. .. deprecated:: 1.8 `n_jobs` is deprecated in version 1.8 and will be removed in 1.10.	None

Estatus de la Salud Mental en adultos del municipio más acaudalado de América Látina.¶

1.1 Introducción.¶

1.2 Objetivo.¶

2.1 Planteamiento.¶

2.2 Contexto de los datos.¶

3.1 Exploración y comprensión del conjunto de datos.¶

4.1 Selección de características.¶

5.1 Definición de variables.¶

6.1 Construcción.¶

Regresión Logística.¶

Linear Discriminant Analysis.¶

Árbol de Decisión.¶

Random Forest Classifier.¶

Support Vector Machine (SVM).¶

Red Neuronal.¶

7.1 Resultados.¶

8.1 Discusiones.¶

9.1 Conclusiones.¶

10.1 Aprendizajes.¶

11.1 Posibles líneas futuras.¶

12.1 Referencias.¶

13.1 Código de Honor de la Universidad de Monterrey.¶

	fecha	id_consulta	edad	sexo	peso	altura	municipio_unidad_medica	institucion_unidad_medica	clave_grupo_ enfermedad	descripcion_grupo_enfermedad	clave_enfermedad	descripcion_enfermedad
0	02/10/2024	SM_2024_38869	21	Masculino	82	174	LINARES	HOSPITAL GENERAL DE LINARES	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F412	TRASTORNO MIXTO DE ANSIEDAD Y DEPRESION
1	08/10/2024	SM_2024_38870	21	Masculino	82	174	LINARES	HOSPITAL GENERAL DE LINARES	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F412	TRASTORNO MIXTO DE ANSIEDAD Y DEPRESION
2	08/10/2024	SM_2024_38871	5	Masculino	21	111	LINARES	HOSPITAL GENERAL DE LINARES	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F919	TRASTORNO DE LA CONDUCTA NO ESPECIFICADO
3	09/10/2024	SM_2024_38872	69	Masculino	sin valor	sin valor	LINARES	HOSPITAL GENERAL DE LINARES	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F321	EPISODIO DEPRESIVO MODERADO
4	09/10/2024	SM_2024_38873	78	Masculino	sin valor	sin valor	LINARES	HOSPITAL GENERAL DE LINARES	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F321	EPISODIO DEPRESIVO MODERADO

	fecha	id_consulta	edad	sexo	peso	altura	municipio_unidad_medica	institucion_unidad_medica	clave_grupo_ enfermedad	descripcion_grupo_enfermedad	clave_enfermedad	descripcion_enfermedad
2625	02/10/2024	SM_2024_41494	33	Masculino	60	160	SAN PEDRO GARZA GARCIA	CENTRO COMUNITARIO DE SALUD MENTAL Y ADICCIONE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F122	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO DEBID...
2626	02/10/2024	SM_2024_41495	53	Femenino	60	160	SAN PEDRO GARZA GARCIA	CENTRO COMUNITARIO DE SALUD MENTAL Y ADICCIONE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F412	TRASTORNO MIXTO DE ANSIEDAD Y DEPRESION
2627	02/10/2024	SM_2024_41496	60	Femenino	60	160	SAN PEDRO GARZA GARCIA	CENTRO COMUNITARIO DE SALUD MENTAL Y ADICCIONE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F630	JUEGO PATOLOGICO
2628	04/10/2024	SM_2024_41497	46	Femenino	70	165	SAN PEDRO GARZA GARCIA	CENTRO COMUNITARIO DE SALUD MENTAL Y ADICCIONE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F412	TRASTORNO MIXTO DE ANSIEDAD Y DEPRESION
2629	04/10/2024	SM_2024_41498	47	Femenino	60	160	SAN PEDRO GARZA GARCIA	CENTRO COMUNITARIO DE SALUD MENTAL Y ADICCIONE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F412	TRASTORNO MIXTO DE ANSIEDAD Y DEPRESION

	fecha	id_consulta	edad	sexo	peso	altura	municipio_unidad_medica	institucion_unidad_medica	clave_grupo_ enfermedad	descripcion_grupo_enfermedad	clave_enfermedad	descripcion_enfermedad	sexo_num	institucion_unidad_medica_num	descripcion_grupo_enfermedad_num	clave_enfermedad_num	descripcion_enfermedad_num
11244	11/12/2024	SM_2024_50113	40	Femenino	86.5	156	SAN PEDRO GARZA GARCIA	CENTRO DE SALUD CON SERVICIOS AMPLIADOS SAN PE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F411	TRASTORNO DE ANSIEDAD GENERALIZADA	0	1	1	31	29
32284	16/05/2025	SM_2025_19420	33	Femenino	57	153	SAN PEDRO GARZA GARCIA	CENTRO DE SALUD CON SERVICIOS AMPLIADOS SAN PE...	XXI	FACTORES QUE INFLUYEN EN EL ESTADO DE SALUD Y ...	Z630	PROBLEMAS EN LA RELACION ENTRE ESPOSOS O PAREJA	0	1	0	46	16
37258	17/06/2025	SM_2025_24394	42	Femenino	46.5	146	SAN PEDRO GARZA GARCIA	CENTRO DE SALUD CON SERVICIOS AMPLIADOS SAN PE...	XXI	FACTORES QUE INFLUYEN EN EL ESTADO DE SALUD Y ...	Z630	PROBLEMAS EN LA RELACION ENTRE ESPOSOS O PAREJA	0	1	0	46	16
41673	22/07/2025	SM_2025_28809	43	Masculino	75	175	SAN PEDRO GARZA GARCIA	CENTRO COMUNITARIO DE SALUD MENTAL Y ADICCIONE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F100	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO DEBID...	1	0	1	0	40
36851	11/06/2025	SM_2025_23987	24	Masculino	83	175	SAN PEDRO GARZA GARCIA	CENTRO COMUNITARIO DE SALUD MENTAL Y ADICCIONE...	V	TRASTORNOS MENTALES Y DEL COMPORTAMIENTO	F412	TRASTORNO MIXTO DE ANSIEDAD Y DEPRESION	1	0	1	32	38

	n_estimators n_estimators: int, default=100 The number of trees in the forest. .. versionchanged:: 0.22 The default value of ``n_estimators`` changed from 10 to 100 in 0.22.	100
	criterion criterion: {"gini", "entropy", "log_loss"}, default="gini" The function to measure the quality of a split. Supported criteria are "gini" for the Gini impurity and "log_loss" and "entropy" both for the Shannon information gain, see :ref:`tree_mathematical_formulation`. Note: This parameter is tree-specific.	'gini'
	max_depth max_depth: int, default=None The maximum depth of the tree. If None, then nodes are expanded until all leaves are pure or until all leaves contain less than min_samples_split samples.	None
	min_samples_split min_samples_split: int or float, default=2 The minimum number of samples required to split an internal node: - If int, then consider `min_samples_split` as the minimum number. - If float, then `min_samples_split` is a fraction and `ceil(min_samples_split * n_samples)` are the minimum number of samples for each split. .. versionchanged:: 0.18 Added float values for fractions.	2
	min_samples_leaf min_samples_leaf: int or float, default=1 The minimum number of samples required to be at a leaf node. A split point at any depth will only be considered if it leaves at least ``min_samples_leaf`` training samples in each of the left and right branches. This may have the effect of smoothing the model, especially in regression. - If int, then consider `min_samples_leaf` as the minimum number. - If float, then `min_samples_leaf` is a fraction and `ceil(min_samples_leaf * n_samples)` are the minimum number of samples for each node. .. versionchanged:: 0.18 Added float values for fractions.	1
	min_weight_fraction_leaf min_weight_fraction_leaf: float, default=0.0 The minimum weighted fraction of the sum total of weights (of all the input samples) required to be at a leaf node. Samples have equal weight when sample_weight is not provided.	0.0
	max_features max_features: {"sqrt", "log2", None}, int or float, default="sqrt" The number of features to consider when looking for the best split: - If int, then consider `max_features` features at each split. - If float, then `max_features` is a fraction and `max(1, int(max_features * n_features_in_))` features are considered at each split. - If "sqrt", then `max_features=sqrt(n_features)`. - If "log2", then `max_features=log2(n_features)`. - If None, then `max_features=n_features`. .. versionchanged:: 1.1 The default of `max_features` changed from `"auto"` to `"sqrt"`. Note: the search for a split does not stop until at least one valid partition of the node samples is found, even if it requires to effectively inspect more than ``max_features`` features.	'sqrt'
	max_leaf_nodes max_leaf_nodes: int, default=None Grow trees with ``max_leaf_nodes`` in best-first fashion. Best nodes are defined as relative reduction in impurity. If None then unlimited number of leaf nodes.	None
	min_impurity_decrease min_impurity_decrease: float, default=0.0 A node will be split if this split induces a decrease of the impurity greater than or equal to this value. The weighted impurity decrease equation is the following:: N_t / N * (impurity - N_t_R / N_t * right_impurity - N_t_L / N_t * left_impurity) where ``N`` is the total number of samples, ``N_t`` is the number of samples at the current node, ``N_t_L`` is the number of samples in the left child, and ``N_t_R`` is the number of samples in the right child. ``N``, ``N_t``, ``N_t_R`` and ``N_t_L`` all refer to the weighted sum, if ``sample_weight`` is passed. .. versionadded:: 0.19	0.0
	bootstrap bootstrap: bool, default=True Whether bootstrap samples are used when building trees. If False, the whole dataset is used to build each tree.	True
	oob_score oob_score: bool or callable, default=False Whether to use out-of-bag samples to estimate the generalization score. By default, :func:`~sklearn.metrics.accuracy_score` is used. Provide a callable with signature `metric(y_true, y_pred)` to use a custom metric. Only available if `bootstrap=True`. For an illustration of out-of-bag (OOB) error estimation, see the example :ref:`sphx_glr_auto_examples_ensemble_plot_ensemble_oob.py`.	False
	n_jobs n_jobs: int, default=None The number of jobs to run in parallel. :meth:`fit`, :meth:`predict`, :meth:`decision_path` and :meth:`apply` are all parallelized over the trees. ``None`` means 1 unless in a :obj:`joblib.parallel_backend` context. ``-1`` means using all processors. See :term:`Glossary ` for more details.	None
	random_state random_state: int, RandomState instance or None, default=None Controls both the randomness of the bootstrapping of the samples used when building trees (if ``bootstrap=True``) and the sampling of the features to consider when looking for the best split at each node (if ``max_features < n_features``). See :term:`Glossary ` for details.	0
	verbose verbose: int, default=0 Controls the verbosity when fitting and predicting.	0
	warm_start warm_start: bool, default=False When set to ``True``, reuse the solution of the previous call to fit and add more estimators to the ensemble, otherwise, just fit a whole new forest. See :term:`Glossary ` and :ref:`tree_ensemble_warm_start` for details.	False
	class_weight class_weight: {"balanced", "balanced_subsample"}, dict or list of dicts, default=None Weights associated with classes in the form ``{class_label: weight}``. If not given, all classes are supposed to have weight one. For multi-output problems, a list of dicts can be provided in the same order as the columns of y. Note that for multioutput (including multilabel) weights should be defined for each class of every column in its own dict. For example, for four-class multilabel classification weights should be [{0: 1, 1: 1}, {0: 1, 1: 5}, {0: 1, 1: 1}, {0: 1, 1: 1}] instead of [{1:1}, {2:5}, {3:1}, {4:1}]. The "balanced" mode uses the values of y to automatically adjust weights inversely proportional to class frequencies in the input data as ``n_samples / (n_classes * np.bincount(y))`` The "balanced_subsample" mode is the same as "balanced" except that weights are computed based on the bootstrap sample for every tree grown. For multi-output, the weights of each column of y will be multiplied. Note that these weights will be multiplied with sample_weight (passed through the fit method) if sample_weight is specified.	None
	ccp_alpha ccp_alpha: non-negative float, default=0.0 Complexity parameter used for Minimal Cost-Complexity Pruning. The subtree with the largest cost complexity that is smaller than ``ccp_alpha`` will be chosen. By default, no pruning is performed. See :ref:`minimal_cost_complexity_pruning` for details. See :ref:`sphx_glr_auto_examples_tree_plot_cost_complexity_pruning.py` for an example of such pruning. .. versionadded:: 0.22	0.0
	max_samples max_samples: int or float, default=None If bootstrap is True, the number of samples to draw from X to train each base estimator. - If None (default), then draw `X.shape[0]` samples. - If int, then draw `max_samples` samples. - If float, then draw `max(round(n_samples * max_samples), 1)` samples. Thus, `max_samples` should be in the interval `(0.0, 1.0]`. .. versionadded:: 0.22	None
	monotonic_cst monotonic_cst: array-like of int of shape (n_features), default=None Indicates the monotonicity constraint to enforce on each feature. - 1: monotonic increase - 0: no constraint - -1: monotonic decrease If monotonic_cst is None, no constraints are applied. Monotonicity constraints are not supported for: - multiclass classifications (i.e. when `n_classes > 2`), - multioutput classifications (i.e. when `n_outputs_ > 1`), - classifications trained on data with missing values. The constraints hold over the probability of the positive class. Read more in the :ref:`User Guide `. .. versionadded:: 1.4	None

	C C: float, default=1.0 Regularization parameter. The strength of the regularization is inversely proportional to C. Must be strictly positive. The penalty is a squared l2 penalty. For an intuitive visualization of the effects of scaling the regularization parameter C, see :ref:`sphx_glr_auto_examples_svm_plot_svm_scale_c.py`.	1.0
	kernel kernel: {'linear', 'poly', 'rbf', 'sigmoid', 'precomputed'} or callable, default='rbf' Specifies the kernel type to be used in the algorithm. If none is given, 'rbf' will be used. If a callable is given it is used to pre-compute the kernel matrix from data matrices; that matrix should be an array of shape ``(n_samples, n_samples)``. For an intuitive visualization of different kernel types see :ref:`sphx_glr_auto_examples_svm_plot_svm_kernels.py`.	'rbf'
	degree degree: int, default=3 Degree of the polynomial kernel function ('poly'). Must be non-negative. Ignored by all other kernels.	3
	gamma gamma: {'scale', 'auto'} or float, default='scale' Kernel coefficient for 'rbf', 'poly' and 'sigmoid'. - if ``gamma='scale'`` (default) is passed then it uses 1 / (n_features * X.var()) as value of gamma, - if 'auto', uses 1 / n_features - if float, must be non-negative. .. versionchanged:: 0.22 The default value of ``gamma`` changed from 'auto' to 'scale'.	'scale'
	coef0 coef0: float, default=0.0 Independent term in kernel function. It is only significant in 'poly' and 'sigmoid'.	0.0
	shrinking shrinking: bool, default=True Whether to use the shrinking heuristic. See the :ref:`User Guide `.	True
	probability probability: bool, default=False Whether to enable probability estimates. This must be enabled prior to calling `fit`, will slow down that method as it internally uses 5-fold cross-validation, and `predict_proba` may be inconsistent with `predict`. Read more in the :ref:`User Guide `.	False
	tol tol: float, default=1e-3 Tolerance for stopping criterion.	0.001
	cache_size cache_size: float, default=200 Specify the size of the kernel cache (in MB).	200
	class_weight class_weight: dict or 'balanced', default=None Set the parameter C of class i to class_weight[i]C for SVC. If not given, all classes are supposed to have weight one. The "balanced" mode uses the values of y to automatically adjust weights inversely proportional to class frequencies in the input data as ``n_samples / (n_classes np.bincount(y))``.	None
	verbose verbose: bool, default=False Enable verbose output. Note that this setting takes advantage of a per-process runtime setting in libsvm that, if enabled, may not work properly in a multithreaded context.	False
	max_iter max_iter: int, default=-1 Hard limit on iterations within solver, or -1 for no limit.	-1
	decision_function_shape decision_function_shape: {'ovo', 'ovr'}, default='ovr' Whether to return a one-vs-rest ('ovr') decision function of shape (n_samples, n_classes) as all other classifiers, or the original one-vs-one ('ovo') decision function of libsvm which has shape (n_samples, n_classes * (n_classes - 1) / 2). However, note that internally, one-vs-one ('ovo') is always used as a multi-class strategy to train models; an ovr matrix is only constructed from the ovo matrix. The parameter is ignored for binary classification. .. versionchanged:: 0.19 decision_function_shape is 'ovr' by default. .. versionadded:: 0.17 decision_function_shape='ovr' is recommended. .. versionchanged:: 0.17 Deprecated decision_function_shape='ovo' and None.	'ovr'
	break_ties break_ties: bool, default=False If true, ``decision_function_shape='ovr'``, and number of classes > 2, :term:`predict` will break ties according to the confidence values of :term:`decision_function`; otherwise the first class among the tied classes is returned. Please note that breaking ties comes at a relatively high computational cost compared to a simple predict. See :ref:`sphx_glr_auto_examples_svm_plot_svm_tie_breaking.py` for an example of its usage with ``decision_function_shape='ovr'``. .. versionadded:: 0.22	False
	random_state random_state: int, RandomState instance or None, default=None Controls the pseudo random number generation for shuffling the data for probability estimates. Ignored when `probability` is False. Pass an int for reproducible output across multiple function calls. See :term:`Glossary `.	0

Modelo	Escenario	Accuracy	G-mean	F1 Clase 0	F1 Clase 1
Regresión Logística	Desbalanceado	81%	59%	50%	89%
Regresión Logística	SMOTE	74%	74%	76%	73%
Árbol de Decisión	Desbalanceado	84%	-	-	-
Árbol de Decisión	SMOTE	81%	-	-	-
Árbol Podado	Desbalanceado	86%	-	-	-
Árbol Podado	SMOTE	79%	-	-	-
Random Forest	Desbalanceado	86%	75%	68%	91%
Random Forest	SMOTE	89%	89%	90%	90%
SVM	Desbalanceado	81%	68%	57%	88%
SVM	SMOTE	75%	75%	77%	74%
Red Neuronal	Desbalanceado	76%	0%	0%	86%
Red Neuronal	SMOTE	74%	74%	76%	73%