Random Forest all data vs subsamples

rf_subsamples.py

import numpy as np
from scipy.stats import chi2
from sklearn.ensemble import RandomForestClassifier

n = 1000
p = 100

def genr_data(n,p):
  X = np.random.randn(n,p)
  y = np.zeros(n)
  for i in range(n):
    y[i] = 1 if np.sum(X[i]**2) > chi2.ppf(0.5,p) else -1
  return (X,y)

d_train = genr_data(n,p)
X_train = d_train[0]
y_train = d_train[1]

d_test = genr_data(10000,p)
X_test = d_test[0]
y_test = d_test[1]

md = RandomForestClassifier(n_estimators = 500, n_jobs = -1)
%time md.fit(X_train, y_train)

yp = md.predict(X_test)
float(np.sum(yp!=y_test))/y_test.size