What does random.sample() method in python do?

Question

I Googled it a lot but could not found it. I want to know the use of random.sample() method and what does it give? When should it be used and some example usage.

Answer 1

According to documentation:

random.sample(population, k)

Return a k length list of unique elements chosen from the population sequence. Used for random sampling without replacement.

Basically, it picks k unique random elements, a sample, from a sequence:

>>> import random
>>> c = list(range(0, 15))
>>> c
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]
>>> random.sample(c, 5)
[9, 2, 3, 14, 11]

random.sample works also directly from a range:

>>> c = range(0, 15)
>>> c
range(0, 15)
>>> random.sample(c, 5)
[12, 3, 6, 14, 10]

In addition to sequences, random.sample works with sets too:

>>> c = {1, 2, 4}
>>> random.sample(c, 2)
[4, 1]

However, random.sample doesn't work with arbitrary iterators:

>>> c = [1, 3]
>>> random.sample(iter(c), 5)
TypeError: Population must be a sequence or set.  For dicts, use list(d).

Answer 2

random.sample() also works on text

example:

> text = open("textfile.txt").read() 

> random.sample(text, 5)

> ['f', 's', 'y', 'v', '\n']

\n is also seen as a character so that can also be returned

you could use random.sample() to return random words from a text file if you first use the split method

example:

> words = text.split()

> random.sample(words, 5)

> ['the', 'and', 'a', 'her', 'of']

Answer 3

random.sample(population, k)

It is used for randomly sampling a sample of length 'k' from a population. returns a 'k' length list of unique elements chosen from the population sequence or set

it returns a new list and leaves the original population unchanged and the resulting list is in selection order so that all sub-slices will also be valid random samples

I am putting up an example in which I am splitting a dataset randomly. It is basically a function in which you pass x_train(population) as an argument and return indices of 60% of the data as D_test.

import random

def randomly_select_70_percent_of_data_from_1_to_length(x_train):
    return random.sample(range(0, len(x_train)), int(0.6*len(x_train)))