Spaces:
Sleeping
Sleeping
import streamlit as st | |
import pandas as pd | |
import numpy as np | |
from sklearn import datasets | |
from sklearn.model_selection import train_test_split | |
from sklearn.tree import DecisionTreeClassifier | |
from sklearn.metrics import accuracy_score | |
import matplotlib.pyplot as plt | |
from sklearn.metrics import confusion_matrix | |
class Decision_tree_st: | |
def __init__(self, database, test_size=0.2): | |
self.database = database | |
self.test_size = test_size | |
self.desc = r''' | |
# **Decision Tree** | |
**Entropy** | |
$$ | |
E = - \sum p(X) \cdot log_{2}(p(X)) | |
$$ | |
$$ | |
p(X) = \frac{len(x)}{n} | |
$$ | |
**Ganancia de información** | |
$$ | |
IG = E(parent) - [weight \quad average] \cdot E(children) | |
$$ | |
**Método (para construir el árbol)** | |
- Se comienza desde el primer nodo y para cada se selecciona la mejor separación en base a la ganancia de información. | |
- De la ganancia de información más alta se rescata la variable y el límite. | |
- Luego se aplica la segmentación a cada nodo, en base a la variable y limite encontrado. | |
- Se itera con estos pasos hasta cumplirse algún criterio | |
- **maximium depth**: cantidad de nodos máximos al final | |
- **minimum samples**: cantidad mínima de elementos que puede tener los nodos | |
- **no more class distribution**: No existen más elementos para segmentar | |
**Aproximación (predicción)** | |
- Se sigue las segmentaciones en el orden del árbol (de arriba a abajo) | |
- Cuando se llega a un nodo al final del árbol se predice según el valor más común en esa muestra. | |
''' | |
self.max_depth = 100 | |
self.min_samples_split = 2 | |
self.stop_criterion = 'max_depth' | |
def params(self): | |
self.stop_criterion = st.radio('Criterio de termino:', options=['max_depth', 'min_samples_split']) | |
if self.stop_criterion == 'max_depth': self.max_depth = st.slider('Valor max deph:', 1, 100, 10) | |
elif self.stop_criterion == 'min_samples_split': self.min_samples_split = st.slider('Valor min_samples_split:', 2, 1000, 5) | |
def solve(self): | |
self.X, self.y = self.database.data, self.database.target | |
X_train, X_test, y_train, y_test = train_test_split(self.X, self.y, test_size=self.test_size, random_state=1234) | |
if self.stop_criterion == 'max_depth': self.sklearn_clf = DecisionTreeClassifier(max_depth=self.max_depth, random_state=1234) | |
elif self.stop_criterion == 'min_samples_split': self.sklearn_clf = DecisionTreeClassifier(min_samples_split=self.min_samples_split, random_state=1234) | |
self.sklearn_clf.fit(X_train, y_train) | |
y_pred = self.sklearn_clf.predict(X_test) | |
acc = accuracy_score(y_pred, y_test) | |
c1, c2 = st.columns([4, 1]) | |
c2.metric('Acierto', value=f'{np.round(acc, 2)*100}%') | |
df = pd.DataFrame(confusion_matrix(y_pred, y_test)) | |
labels = self.database.target_names | |
df.columns = labels | |
df.index = labels | |
c1.write('**Confusion Matrix**') | |
c1.dataframe(df) | |
def visualization(self): | |
n_features = int(self.database.data.shape[1]) | |
self.x_feature = st.slider('Variables en eje x', 1, n_features, 1) | |
self.y_feature = st.slider('Variables en eje y', 1, n_features, 2) | |
self.X = np.c_[self.database.data[:, self.x_feature-1:self.x_feature], self.database.data[:, self.y_feature-1:self.y_feature]] | |
self.y = self.database.target | |
X_train, X_test, y_train, y_test = train_test_split(self.X, self.y, test_size=self.test_size, random_state=1234) | |
if self.stop_criterion == 'max_depth': self.sklearn_clf = DecisionTreeClassifier(max_depth=self.max_depth, random_state=1234) | |
elif self.stop_criterion == 'min_samples_split': self.sklearn_clf = DecisionTreeClassifier(min_samples_split=self.min_samples_split, random_state=1234) | |
self.sklearn_clf.fit(X_train, y_train) | |
x1_min, x1_max = self.X[:, 0].min() - 0.5, self.X[:, 0].max() + 0.5 | |
x2_min, x2_max = self.X[:, 1].min() - 0.5, self.X[:, 1].max() + 0.5 | |
h = 0.02 # Salto que vamos dando | |
x1_i = np.arange(x1_min, x1_max, h) | |
x2_i = np.arange(x2_min, x2_max, h) | |
x1_x1, x2_x2 = np.meshgrid(x1_i, x2_i) | |
y_pred = self.sklearn_clf.predict(np.c_[x1_x1.ravel(), x2_x2.ravel()]) | |
y_pred = y_pred.reshape(x1_x1.shape) | |
plt.figure(1, figsize=(12, 8)) | |
plt.pcolormesh(x1_x1, x2_x2, y_pred, cmap=plt.cm.Paired) | |
plt.scatter(self.X[:, 0], self.X[:, 1], c=self.y, edgecolors='k', cmap=plt.cm.Paired) | |
plt.xlim(x1_x1.min(), x1_x1.max()) | |
plt.ylim(x2_x2.min(), x2_x2.max()) | |
return plt.gcf() | |