CODEWITHSUNDEEP
pythonpandasnumpyrandomrandom-seed
Martin Thoma
Martin Thoma

Reputation: 136359

Does Pandas use Numpy as a random number generator?

I want to get reproducible samples of data. A quick experiment suggests, that numpy.random.seed does influence pandas.DataFrame.sample, but it is not documented.

Does anybody know

What I tried

I ran the following a couple of times and always got the same results back

#!/usr/bin/env python

import pandas as pd
import numpy as np


df = pd.DataFrame([(1, 2, 1),
                   (1, 2, 2),
                   (1, 2, 3),
                   (4, 1, 612),
                   (4, 1, 612),
                   (4, 1, 1),
                   (3, 2, 1),
                   ],
                  columns=['groupid', 'a', 'b'],
                  index=['India', 'France', 'England', 'Germany', 'UK', 'USA',
                         'Indonesia'])
np.random.seed(0)
print(df.sample(n=1))
print(df.sample(n=1))
print(df.sample(n=1))
print(df.sample(n=1))
print(df.sample(n=1))

Which gives:

Upvotes: 2

Views: 1551

Answers (1)

jezrael
jezrael

Reputation: 862661

pandas use _random_state function for return np.random.RandomState link:

def _random_state(state=None):
    """
    Helper function for processing random_state arguments.
    Parameters
    ----------
    state : int, np.random.RandomState, None.
        If receives an int, passes to np.random.RandomState() as seed.
        If receives an np.random.RandomState object, just returns object.
        If receives `None`, returns np.random.
        If receives anything else, raises an informative ValueError.
        Default None.
    Returns
    -------
    np.random.RandomState
    """

    if types.is_integer(state):
        return np.random.RandomState(state)
    elif isinstance(state, np.random.RandomState):
        return state
    elif state is None:
        return np.random
    else:
        raise ValueError("random_state must be an integer, a numpy "
                         "RandomState, or None")

and in sample is called this function:

    # Process random_state argument
    rs = com._random_state(random_state)

Upvotes: 1

Related Questions