Ejemplo árboles de decisión empresas en re organización#
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.ticker as mtick
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
from sklearn import tree
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import classification_report, confusion_matrix, roc_auc_score, roc_curve
from sklearn.metrics import ConfusionMatrixDisplay, precision_score, precision_recall_curve, recall_score, accuracy_score, f1_score
# path = "BD empresas re organización.xlsx"
path = "BD empresas en re organización.xlsx"
xls = pd.ExcelFile(path)
df = pd.read_excel(path, sheet_name=xls.sheet_names[0])
df.head()
| Razón Social | Margen EBIT | Carga financiera | Margen neto | CxC | CxP | Solvencia | Apalancamiento | En Reorganización | |
|---|---|---|---|---|---|---|---|---|---|
| 0 | AACER SAS | 0.071690 | 0.000000 | 0.042876 | 0.104095 | 0.153192 | 1.877078 | 1.642505 | 0 |
| 1 | ABARROTES EL ROMPOY SAS | 0.017816 | 0.000000 | 0.010767 | 0.018414 | 0.000000 | 0.000000 | 0.865044 | 0 |
| 2 | ABASTECIMIENTOS INDUSTRIALES SAS | 0.144646 | 0.054226 | 0.059784 | 0.227215 | 0.025591 | 1.077412 | 1.272299 | 0 |
| 3 | ACME LEON PLASTICOS SAS | 0.004465 | 0.000000 | -0.013995 | 0.073186 | 0.127866 | 0.000000 | 1.391645 | 0 |
| 4 | ADVANCED PRODUCTS COLOMBIA SAS | 0.141829 | 0.050810 | 0.053776 | 0.398755 | 0.147678 | 0.675073 | 2.118774 | 0 |
# Conteo absoluto
conteo_clases = df['En Reorganización'].value_counts()
# Porcentaje
porcentaje_clases = df['En Reorganización'].value_counts(normalize=True) * 100
# Mostrar conteo y porcentaje
print("Cantidad de empresas por clase:")
print(conteo_clases)
print("\nPorcentaje de empresas por clase:")
print(porcentaje_clases.round(2))
Cantidad de empresas por clase:
En Reorganización
1 342
0 287
Name: count, dtype: int64
Porcentaje de empresas por clase:
En Reorganización
1 54.37
0 45.63
Name: proportion, dtype: float64
Árboles de decisión:#
# ------------------------
# Selección de variables
# ------------------------
variables_seleccionadas = ['Margen EBIT',
'Carga financiera',
'Margen neto',
'CxC',
'CxP',
'Solvencia',
'Apalancamiento']
# Variable objetivo
target = 'En Reorganización'
# ------------------------
# Preparar datos
# ------------------------
X = df[variables_seleccionadas]
y = df[target]
# Estandarizar variables
# scaler = StandardScaler()
# X_scaled = scaler.fit_transform(X)
# Dividir en entrenamiento y prueba (70%-30%)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=35, stratify=y)
stratify=y le dice a train_test_split que mantenga la misma
proporción de clases de y (variable objetivo) en los subconjuntos de
train y test.
# ------------------------
# Ajustar el modelo
# ------------------------
model = DecisionTreeClassifier(
max_depth=None, min_samples_split=10, min_samples_leaf=2, random_state=34
)
model.fit(X_train, y_train)
# ------------------------
# Predicciones
# ------------------------
y_pred_train = model.predict(X_train)
y_prob_train = model.predict_proba(X_train)[:, 1]
y_pred = model.predict(X_test)
y_prob = model.predict_proba(X_test)[:, 1]
# ------------------------
# Evaluación del modelo
# ------------------------
cm_train = confusion_matrix(y_train, y_pred_train, labels=[0,1])
cm_df_train = pd.DataFrame(cm_train, index=["Real 0", "Real 1"], columns=["Predicho 0", "Predicho 1"])
plt.figure(figsize=(5.2,4.2))
sns.heatmap(cm_train, annot=True, fmt="d", cbar=True, linewidths=.5, cmap="coolwarm")
plt.title("Matriz de confusión - train")
plt.xlabel("Predicho"); plt.ylabel("Real")
plt.tight_layout()
plt.show()
cm = confusion_matrix(y_test, y_pred, labels=[0,1])
cm_df = pd.DataFrame(cm, index=["Real 0", "Real 1"], columns=["Predicho 0", "Predicho 1"])
plt.figure(figsize=(5.2,4.2))
sns.heatmap(cm_df, annot=True, fmt="d", cbar=True, linewidths=.5, cmap="coolwarm")
plt.title("Matriz de confusión - Test")
plt.xlabel("Predicho"); plt.ylabel("Real")
plt.tight_layout()
plt.show()
print("\n=== Reporte de Clasificación - train ===")
print(classification_report(y_train, y_pred_train))
print("\n=== Reporte de Clasificación - test ===")
print(classification_report(y_test, y_pred))
=== Reporte de Clasificación - train ===
precision recall f1-score support
0 0.93 0.93 0.93 201
1 0.94 0.94 0.94 239
accuracy 0.93 440
macro avg 0.93 0.93 0.93 440
weighted avg 0.93 0.93 0.93 440
=== Reporte de Clasificación - test ===
precision recall f1-score support
0 0.83 0.80 0.82 86
1 0.84 0.86 0.85 103
accuracy 0.84 189
macro avg 0.84 0.83 0.83 189
weighted avg 0.84 0.84 0.84 189
# ============================
# ROC AUC Score
# ============================
auc_train = roc_auc_score(y_train, y_prob_train)
auc_test = roc_auc_score(y_test, y_prob)
print(f"ROC AUC - Train: {auc_train:.3f}")
print(f"ROC AUC - Test : {auc_test:.3f}")
# ============================
# Curva ROC (Train y Test)
# ============================
fpr_train, tpr_train, _ = roc_curve(y_train, y_prob_train)
fpr_test, tpr_test, _ = roc_curve(y_test, y_prob)
plt.figure(figsize=(8, 6))
plt.plot(fpr_train, tpr_train, label=f'Train (AUC = {auc_train:.2f})', color='blue')
plt.plot(fpr_test, tpr_test, label=f'Test (AUC = {auc_test:.2f})', color='orange')
plt.plot([0, 1], [0, 1], 'k--', label='Azar')
plt.xlabel("False Positive Rate")
plt.ylabel("True Positive Rate")
plt.title("Curva ROC - Train y Test")
plt.legend(loc="lower right")
plt.grid(True)
plt.tight_layout()
plt.show()
ROC AUC - Train: 0.989
ROC AUC - Test : 0.883
# Calcular precisión y recall para diferentes umbrales
precision, recall, thresholds = precision_recall_curve(y_test, y_prob)
# Agregar el umbral 0 para completar el array de thresholds
thresholds = np.append(thresholds, 1)
# Graficar precisión y recall en función del umbral
plt.figure(figsize=(10, 6))
plt.plot(thresholds, precision, label="Precisión")
plt.plot(thresholds, recall, label="Recall")
plt.xlabel("Umbral")
plt.ylabel("Precisión/Recall")
plt.title("Precisión y Recall en función del umbral")
plt.legend()
plt.grid(True)
plt.show()
plt.figure(figsize=(8, 6))
plt.plot(recall, precision, marker=".", label="Regresión Logística")
plt.xlabel("Recall")
plt.ylabel("Precisión")
plt.title("Curva de Precisión-Recall")
plt.legend()
plt.grid(True)
plt.show()
y_prob
array([0.57142857, 0. , 0.5 , 0.5 , 0. ,
0.66666667, 0. , 0. , 0. , 1. ,
0. , 0.66666667, 0. , 0. , 1. ,
0.5 , 0. , 1. , 0.16666667, 0. ,
0.8 , 1. , 0. , 0. , 1. ,
1. , 1. , 1. , 0.57142857, 1. ,
0. , 0. , 1. , 1. , 1. ,
1. , 0. , 1. , 1. , 0.57142857,
1. , 0.6 , 1. , 0. , 0. ,
1. , 1. , 0. , 0.57142857, 0. ,
1. , 0.25 , 0. , 1. , 1. ,
0. , 0. , 1. , 0. , 0. ,
0. , 1. , 0. , 1. , 0. ,
1. , 0.6 , 0.2 , 0.57142857, 1. ,
0. , 0. , 0.66666667, 0.25 , 0.2 ,
1. , 0. , 0.66666667, 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
1. , 1. , 0. , 1. , 1. ,
1. , 0. , 0.66666667, 1. , 1. ,
1. , 0. , 1. , 0. , 1. ,
0. , 1. , 1. , 0.66666667, 1. ,
1. , 0. , 0. , 1. , 1. ,
0. , 0.66666667, 0. , 1. , 0. ,
0.2 , 1. , 0. , 0. , 0. ,
1. , 1. , 1. , 1. , 1. ,
0.66666667, 0. , 0.66666667, 0. , 0. ,
0. , 1. , 1. , 1. , 0.66666667,
1. , 1. , 0.6 , 0.66666667, 1. ,
0. , 0.5 , 0. , 1. , 1. ,
1. , 1. , 1. , 0.57142857, 0. ,
1. , 0.6 , 1. , 1. , 0. ,
0.5 , 0. , 1. , 1. , 1. ,
1. , 0.5 , 0. , 0.66666667, 1. ,
1. , 0.5 , 1. , 0.57142857, 0. ,
1. , 0. , 1. , 0.5 , 1. ,
0. , 0.5 , 0. , 1. , 0.8 ,
0.66666667, 0. , 1. , 1. , 0. ,
1. , 0. , 0. , 0.6 ])
# DataFrame con probas y clase real
df_deciles = pd.DataFrame({'y_real': y_test, 'y_proba': y_prob})
# Crear deciles (1 = más alto riesgo, 10 = más bajo)
df_deciles['Decil'] = pd.qcut(df_deciles['y_proba'], 10, labels=False, duplicates='drop') + 1
df_deciles['Decil'] = 11 - df_deciles['Decil'] # invertir para que el decil 1 sea el de mayor riesgo
# Calcular tasa por decil
tabla_deciles = df_deciles.groupby('Decil').agg(
Total=('y_real','count'),
Positivos=('y_real','sum')
)
tabla_deciles['Tasa'] = tabla_deciles['Positivos'] / tabla_deciles['Total']
tabla_deciles['Lift'] = tabla_deciles['Tasa'] / df_deciles['y_real'].mean()
tabla_deciles['Captura_Acum'] = tabla_deciles['Positivos'].cumsum() / df_deciles['y_real'].sum()
print(f"Tasa de positivos reales en test: {df_deciles['y_real'].mean():.2f}")
print(tabla_deciles)
# --- 📊 Gráfico ---
plt.figure(figsize=(8,5))
plt.plot(tabla_deciles.index, tabla_deciles['Tasa'], marker='o', linestyle='-', color='blue')
plt.title("Tasa de positivos por decil")
plt.xlabel("Decil")
plt.ylabel("Tasa de clase 1")
plt.grid(True)
plt.show()
Tasa de positivos reales en test: 0.54
Total Positivos Tasa Lift Captura_Acum
Decil
8 94 80 0.851064 1.561661 0.776699
9 12 9 0.750000 1.376214 0.864078
10 83 14 0.168675 0.309510 1.000000
Cambio de umbral:#
# Crear lista de umbrales a evaluar
umbrales = np.arange(0.1, 0.91, 0.05)
# Lista para almacenar resultados
resultados = []
for umbral in umbrales:
y_pred_umbral = (y_prob >= umbral).astype(int)
tn, fp, fn, tp = confusion_matrix(y_test, y_pred_umbral).ravel()
precision = precision_score(y_test, y_pred_umbral, zero_division=0)
recall = recall_score(y_test, y_pred_umbral)
specificity = tn / (tn + fp)
accuracy = accuracy_score(y_test, y_pred_umbral)
f1 = f1_score(y_test, y_pred_umbral)
resultados.append({
'Umbral': umbral,
'Precision': precision,
'Recall (Sensibilidad)': recall,
'Especificidad': specificity,
'Accuracy': accuracy,
'F1-score': f1
})
# Convertir a DataFrame
df_resultados = pd.DataFrame(resultados)
# Mostrar tabla
plt.figure(figsize=(12, 6))
sns.lineplot(data=df_resultados.set_index('Umbral'))
plt.title('Métricas por Umbral de Decisión')
plt.ylabel('Valor')
plt.gca().yaxis.set_major_formatter(mtick.PercentFormatter(1.0))
plt.grid(True)
plt.show()
df_resultados
| Umbral | Precision | Recall (Sensibilidad) | Especificidad | Accuracy | F1-score | |
|---|---|---|---|---|---|---|
| 0 | 0.10 | 0.776860 | 0.912621 | 0.686047 | 0.809524 | 0.839286 |
| 1 | 0.15 | 0.776860 | 0.912621 | 0.686047 | 0.809524 | 0.839286 |
| 2 | 0.20 | 0.786325 | 0.893204 | 0.709302 | 0.809524 | 0.836364 |
| 3 | 0.25 | 0.800000 | 0.893204 | 0.732558 | 0.820106 | 0.844037 |
| 4 | 0.30 | 0.800000 | 0.893204 | 0.732558 | 0.820106 | 0.844037 |
| 5 | 0.35 | 0.800000 | 0.893204 | 0.732558 | 0.820106 | 0.844037 |
| 6 | 0.40 | 0.800000 | 0.893204 | 0.732558 | 0.820106 | 0.844037 |
| 7 | 0.45 | 0.800000 | 0.893204 | 0.732558 | 0.820106 | 0.844037 |
| 8 | 0.50 | 0.839623 | 0.864078 | 0.802326 | 0.835979 | 0.851675 |
| 9 | 0.55 | 0.839623 | 0.864078 | 0.802326 | 0.835979 | 0.851675 |
| 10 | 0.60 | 0.851064 | 0.776699 | 0.837209 | 0.804233 | 0.812183 |
| 11 | 0.65 | 0.851064 | 0.776699 | 0.837209 | 0.804233 | 0.812183 |
| 12 | 0.70 | 0.913580 | 0.718447 | 0.918605 | 0.809524 | 0.804348 |
| 13 | 0.75 | 0.913580 | 0.718447 | 0.918605 | 0.809524 | 0.804348 |
| 14 | 0.80 | 0.924051 | 0.708738 | 0.930233 | 0.809524 | 0.802198 |
| 15 | 0.85 | 0.924051 | 0.708738 | 0.930233 | 0.809524 | 0.802198 |
| 16 | 0.90 | 0.924051 | 0.708738 | 0.930233 | 0.809524 | 0.802198 |
umbral_optimo = 0.55
y_pred_final = (y_prob >= umbral_optimo).astype(int)
cm_df_final = confusion_matrix(y_test, y_pred_final)
plt.figure(figsize=(5.2,4.2))
sns.heatmap(cm_df_final, annot=True, fmt="d", cbar=True, linewidths=.5, cmap="coolwarm")
plt.title("Matriz de confusión - Test")
plt.xlabel("Predicho"); plt.ylabel("Real")
plt.tight_layout()
plt.show()
print("\nReporte de Clasificación:")
print(classification_report(y_test, y_pred_final))
print(f"ROC AUC: {roc_auc_score(y_test, y_prob):.3f}")
Reporte de Clasificación:
precision recall f1-score support
0 0.83 0.80 0.82 86
1 0.84 0.86 0.85 103
accuracy 0.84 189
macro avg 0.84 0.83 0.83 189
weighted avg 0.84 0.84 0.84 189
ROC AUC: 0.883
# DataFrame con probas y clase real
df_deciles = pd.DataFrame({'y_real': y_pred_final, 'y_proba': y_prob})
# Crear deciles (1 = más alto riesgo, 10 = más bajo)
df_deciles['Decil'] = pd.qcut(df_deciles['y_proba'], 10, labels=False, duplicates='drop') + 1
df_deciles['Decil'] = 11 - df_deciles['Decil'] # invertir para que el decil 1 sea el de mayor riesgo
# Calcular tasa por decil
tabla_deciles = df_deciles.groupby('Decil').agg(
Total=('y_real','count'),
Positivos=('y_real','sum')
)
tabla_deciles['Tasa'] = tabla_deciles['Positivos'] / tabla_deciles['Total']
tabla_deciles['Lift'] = tabla_deciles['Tasa'] / df_deciles['y_real'].mean()
tabla_deciles['Captura_Acum'] = tabla_deciles['Positivos'].cumsum() / df_deciles['y_real'].sum()
print(f"Tasa de positivos reales en test: {df_deciles['y_real'].mean():.2f}")
print(tabla_deciles)
# --- 📊 Gráfico ---
plt.figure(figsize=(8,5))
plt.plot(tabla_deciles.index, tabla_deciles['Tasa'], marker='o', linestyle='-', color='blue')
plt.title("Tasa de positivos por decil")
plt.xlabel("Decil")
plt.ylabel("Tasa de clase 1")
plt.grid(True)
plt.show()
Tasa de positivos reales en test: 0.56
Total Positivos Tasa Lift Captura_Acum
Decil
8 94 94 1.0 1.783019 0.886792
9 12 12 1.0 1.783019 1.000000
10 83 0 0.0 0.000000 1.000000
