#!/usr/bin/env python
# coding: utf-8

# In[1]:


import numpy as np
from math import ceil, floor
from sklearn.datasets import load_boston
from sklearn.model_selection import ShuffleSplit as skShuffleSplit


# In[2]:


class ShuffleSplit():
    def __init__(self, n_splits=10,
                 train_size=0.9, test_size=0.1, random_state=0):
        self.n_splits = n_splits
        self.train_size = train_size
        self.test_size = test_size
        self.random_state = random_state

    def split(self, X, y):
        n_train = floor(self.train_size * X.shape[0])
        n_test = ceil(self.test_size * X.shape[0])
        rng = np.random.RandomState(self.random_state)
        for _ in range(self.n_splits):
            permutation = rng.permutation(X.shape[0])
            yield (permutation[n_test:(n_test + n_train)],
                   permutation[:n_test])


# In[3]:


X, y = load_boston(return_X_y=True)
cv1 = ShuffleSplit(n_splits=5, random_state=0)
cv2 = skShuffleSplit(n_splits=5, random_state=0)
for (train1, test1), (train2, test2) in zip(cv1.split(X, y), cv2.split(X, y)):
    assert np.array_equal(train1, train2)
    assert np.array_equal(test1, test2)


# In[4]:


X, y = load_boston(return_X_y=True)
cv1 = ShuffleSplit(n_splits=5, train_size=0.5, test_size=0.2, random_state=0)
cv2 = skShuffleSplit(n_splits=5, train_size=0.5, test_size=0.2, random_state=0)
for (train1, test1), (train2, test2) in zip(cv1.split(X, y), cv2.split(X, y)):
    assert np.array_equal(train1, train2)
    assert np.array_equal(test1, test2)