def test_read_csv_sync(loop):
dd = pytest.importorskip('dask.dataframe')
with cluster() as (s, [a, b]):
with Executor(('127.0.0.1', s['port']), loop=loop) as e:
df = read_csv('distributed-test/csv/2015/', lazy=True)
assert isinstance(df, dd.DataFrame)
assert list(df.columns) == ['name', 'amount', 'id']
f = e.compute(df.amount.sum())
assert f.result() == (100 + 200 + 300 + 400 + 500 + 600)
df = read_csv('distributed-test/csv/2015/', lazy=False)
assert df.amount.sum().compute() == (100 + 200 + 300 + 400 + 500 + 600)
def test_read_csv_sync(loop):
dd = pytest.importorskip('dask.dataframe')
with cluster() as (s, [a, b]):
with Executor(('127.0.0.1', s['port']), loop=loop) as e:
df = read_csv('distributed-test/csv/2015/', lazy=True)
assert isinstance(df, dd.DataFrame)
assert list(df.columns) == ['name', 'amount', 'id']
f = e.compute(df.amount.sum())
assert f.result() == (100 + 200 + 300 + 400 + 500 + 600)
df = read_csv('distributed-test/csv/2015/', lazy=False)
assert df.amount.sum().compute() == (100 + 200 + 300 + 400 + 500 + 600)
def test_read_csv_gzip(e, s, a, b):
dd = pytest.importorskip('dask.dataframe')
s3 = S3FileSystem(anon=True)
df = read_csv('distributed-test/csv/gzip/', compression='gzip')
assert isinstance(df, dd.DataFrame)
assert list(df.columns) == ['name', 'amount', 'id']
f = e.compute(df.amount.sum())
result = yield f._result()
assert result == (100 + 200 + 300 + 400 + 500 + 600)
def test_read_csv_gzip(e, s, a, b):
dd = pytest.importorskip('dask.dataframe')
s3 = S3FileSystem(anon=True)
df = read_csv('distributed-test/csv/gzip/', compression='gzip')
assert isinstance(df, dd.DataFrame)
assert list(df.columns) == ['name', 'amount', 'id']
f = e.compute(df.amount.sum())
result = yield f._result()
assert result == (100 + 200 + 300 + 400 + 500 + 600)
def test_read_csv(e, s, a, b):
dd = pytest.importorskip('dask.dataframe')
s3 = S3FileSystem(anon=True)
df = read_csv('distributed-test/csv/2015/*',
lazy=True,
storage_options={'anon': True})
yield gen.sleep(0.1)
assert not s.tasks
assert isinstance(df, dd.DataFrame)
df = read_csv('distributed-test/csv/2015/*',
storage_options={'anon': True})
assert isinstance(df, dd.DataFrame)
assert list(df.columns) == ['name', 'amount', 'id']
f = e.compute(df.amount.sum())
result = yield f._result()
assert result == (100 + 200 + 300 + 400 + 500 + 600)
futures = read_csv('distributed-test/csv/2015/*',
collection=False,
lazy=False,
storage_options={'anon': True})
assert len(futures) == 3
assert all(isinstance(f, Future) for f in futures)
results = yield e._gather(futures)
assert results[0].id.sum() == 1 + 2 + 3
assert results[1].id.sum() == 0
assert results[2].id.sum() == 4 + 5 + 6
values = read_csv('distributed-test/csv/2015/*',
collection=False,
lazy=True,
storage_options={'anon': True})
assert len(values) == 3
assert all(isinstance(v, Delayed) for v in values)
df2 = read_csv('distributed-test/csv/2015/*',
collection=True,
lazy=True,
blocksize=20,
storage_options={'anon': True})
assert df2.npartitions > df.npartitions
result = yield e.compute(df2.id.sum())._result()
assert result == 1 + 2 + 3 + 4 + 5 + 6
df2 = read_csv('distributed-test/csv/2015/*',
collection=True,
lazy=False,
blocksize=20,
storage_options={'anon': True})
f = e.compute(df2.amount.sum())
result = yield f._result()
assert result == (100 + 200 + 300 + 400 + 500 + 600)
def test_read_csv(e, s, a, b):
dd = pytest.importorskip('dask.dataframe')
s3 = S3FileSystem(anon=True)
df = read_csv('distributed-test/csv/2015/*', lazy=True,
storage_options={'anon': True})
yield gen.sleep(0.1)
assert not s.tasks
assert isinstance(df, dd.DataFrame)
df = read_csv('distributed-test/csv/2015/*', storage_options={'anon': True})
assert isinstance(df, dd.DataFrame)
assert list(df.columns) == ['name', 'amount', 'id']
f = e.compute(df.amount.sum())
result = yield f._result()
assert result == (100 + 200 + 300 + 400 + 500 + 600)
futures = read_csv('distributed-test/csv/2015/*',
collection=False, lazy=False,
storage_options={'anon': True})
assert len(futures) == 3
assert all(isinstance(f, Future) for f in futures)
results = yield e._gather(futures)
assert results[0].id.sum() == 1 + 2 + 3
assert results[1].id.sum() == 0
assert results[2].id.sum() == 4 + 5 + 6
values = read_csv('distributed-test/csv/2015/*',
collection=False, lazy=True,
storage_options={'anon': True})
assert len(values) == 3
assert all(isinstance(v, Delayed) for v in values)
df2 = read_csv('distributed-test/csv/2015/*',
collection=True, lazy=True, blocksize=20,
storage_options={'anon': True})
assert df2.npartitions > df.npartitions
result = yield e.compute(df2.id.sum())._result()
assert result == 1 + 2 + 3 + 4 + 5 + 6
df2 = read_csv('distributed-test/csv/2015/*',
collection=True, lazy=False, blocksize=20,
storage_options={'anon': True})
f = e.compute(df2.amount.sum())
result = yield f._result()
assert result == (100 + 200 + 300 + 400 + 500 + 600)