def test_seed():
src = generate_csv(100000)
no_seed_a = list(sample(src, 'random', sample_rate=0.1))
no_seed_b = list(sample(src, 'random', sample_rate=0.1))
seed0_a = list(sample(src, 'random', sample_rate=0.1, seed=0))
seed0_b = list(sample(src, 'random', sample_rate=0.1, seed=0))
seed1 = list(sample(src, 'random', sample_rate=0.1, seed=1))
assert no_seed_a != no_seed_b
assert seed0_a == seed0_b
assert seed0_a != seed1
def test_random_sample():
src = generate_csv(100000)
sample_rate = 0.1
sampled = list(sample(src, 'random', sample_rate=sample_rate))[1:]
# Check sample size
assert abs(len(sampled) / len(src) - sample_rate) < 0.01
# Check elements
assert set(src).issuperset(sampled)
def test_reservoir_sample():
src = generate_csv(100000)
sample_size = 10000
sampled = list(sample(src, 'reservoir', sample_size=sample_size))[1:]
# Check sample size
assert sample_size == len(sampled)
# Check elements
assert set(src).issuperset(sampled)
def test_hash_sample():
src = generate_csv(100000)
sample_rate = 0.1
sampled = list(sample(src, 'hash', sample_rate=sample_rate, col='a'))[1:]
# Check sample size
assert abs(len(sampled) / len(src) - sample_rate) < 0.01
# Check elements
assert set(src).issuperset(sampled)
# Check consistency
assert all(v == 10
for v in Counter(row.split(',')[0] for row in sampled).values())
def test_to_buf():
src = generate_csv(100000)
buf = sample(src, 'random', sample_rate=0.1).to_buf()
assert src[0] + '\n' == buf.readline()
def test_sampling_rate_one():
src = generate_csv(100000)
assert src == list(sample(src, 'random', sample_rate=1))
assert src == list(sample(src, 'hash', sample_rate=1, col='a'))
assert src == list(sample(src, 'reservoir', sample_size=len(src) - 1))
def test_sampling_rate_zero():
src = generate_csv(100000)
assert 0 == len(list(sample(src, 'random', sample_rate=0))[1:])
assert 0 == len(list(sample(src, 'hash', sample_rate=0, col='a'))[1:])
assert 0 == len(list(sample(src, 'reservoir', sample_size=0))[1:])