unexpected keyword argument 'random_state'

[ivanfig]

TypeError Traceback (most recent call last)
in
13 p_h2 = h2.pct_change(1)
14 #split the embedded time series
---> 15 x1tr, x1te, x2tr, x2te = train_test_split(p_h1,p_h2, percent=.85, random_state=rng)
16
17 print(str(len(x1tr)), str(len(x2tr)))

TypeError: train_test_split() got an unexpected keyword argument 'random_state'

[Anmwai]

Seriously, after taking forever to load, I was expecting a solution not just a description of the problem.

[Flashness123]

use seed instead of random_state:
train_test = toy_dataset.train_test_split(test_size=0.05, seed = 42)