def test_header_disagrees_with_dshape():
ds = datashape.dshape("var * {name: string, bal: int64}")
with filetext("name,val\nAlice,100\nBob,200", extension="csv") as fn:
csv = CSV(fn, header=True)
assert convert(list, csv) == [("Alice", 100), ("Bob", 200)]
assert list(convert(pd.DataFrame, csv).columns) == ["name", "val"]
assert list(convert(pd.DataFrame, csv, dshape=ds).columns) == ["name", "bal"]
def test_header_disagrees_with_dshape():
ds = datashape.dshape('var * {name: string, bal: int64}')
with filetext('name,val\nAlice,100\nBob,200', extension='csv') as fn:
csv = CSV(fn, header=True)
assert convert(list, csv) == [('Alice', 100), ('Bob', 200)]
assert list(convert(pd.DataFrame, csv).columns) == ['name', 'val']
assert list(convert(pd.DataFrame, csv, dshape=ds).columns) == ['name', 'bal']
def test_header_mix_str_digits():
ds = datashape.dshape('''var * {"On- or Off- Budget": ?string,
"1990": ?string}''')
with filetext('On- or Off- Budget,1990\nOn Budget,-628\nOff budget,"5,962"\n') as fn:
csv = CSV(fn, has_header=True)
df = convert(pd.DataFrame, csv)
assert discover(csv).measure == ds.measure
def test_pandas_read_supports_read_csv_kwargs():
with filetext("Alice,1\nBob,2") as fn:
ds = datashape.dshape("var * {name: string, amount: int}")
csv = CSV(fn)
df = csv_to_dataframe(csv, dshape=ds, usecols=["name"])
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [("Alice",), ("Bob",)]
def test_pandas_read_supports_read_csv_kwargs():
with filetext('Alice,1\nBob,2') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_dataframe(csv, dshape=ds, usecols=['name'])
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice',), ('Bob',)]
def test_pandas_read_supports_gzip():
with filetext("Alice,1\nBob,2", open=gzip.open, mode="wt", extension=".csv.gz") as fn:
ds = datashape.dshape("var * {name: string, amount: int}")
csv = CSV(fn)
df = csv_to_dataframe(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [("Alice", 1), ("Bob", 2)]
assert list(df.columns) == ["name", "amount"]
def test_pandas_read():
with filetext("Alice,1\nBob,2") as fn:
ds = datashape.dshape("var * {name: string, amount: int}")
csv = CSV(fn)
df = csv_to_dataframe(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [("Alice", 1), ("Bob", 2)]
assert list(df.columns) == ["name", "amount"]
def test_glob():
d = {"accounts1.csv": "name,when\nAlice,100\nBob,200", "accounts2.csv": "name,when\nAlice,300\nBob,400"}
with filetexts(d) as fns:
r = resource("accounts*.csv", has_header=True)
assert convert(list, r) == [("Alice", 100), ("Bob", 200), ("Alice", 300), ("Bob", 400)]
r = resource("*.csv")
assert isinstance(r, chunks(CSV))
def test_unused_datetime_columns():
ds = datashape.dshape('var * {val: string, when: datetime}')
with filetext("val,when\na,2000-01-01\nb,2000-02-02") as fn:
csv = CSV(fn, has_header=True)
assert convert(
list,
csv_to_dataframe(csv, usecols=['val'], squeeze=True,
dshape=ds)) == ['a', 'b']
def test_pandas_loads_in_datetimes_naively():
with filetext('name,when\nAlice,2014-01-01\nBob,2014-02-02') as fn:
csv = CSV(fn, has_header=True)
ds = datashape.dshape('var * {name: ?string, when: ?datetime}')
assert discover(csv) == ds
df = convert(pd.DataFrame, csv)
assert df.dtypes['when'] == 'M8[ns]'
def test_pandas_read():
with filetext('Alice,1\nBob,2') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_dataframe(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice', 1), ('Bob', 2)]
assert list(df.columns) == ['name', 'amount']
def test_pandas_read():
with filetext('Alice,1\nBob,2') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_dataframe(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice', 1), ('Bob', 2)]
assert list(df.columns) == ['name', 'amount']
def test_pandas_loads_in_datetimes_naively():
with filetext('name,when\nAlice,2014-01-01\nBob,2014-02-02') as fn:
csv = CSV(fn, has_header=True)
ds = datashape.dshape('var * {name: ?string, when: ?datetime}')
assert discover(csv) == ds
df = convert(pd.DataFrame, csv)
assert df.dtypes['when'] == 'M8[ns]'
def test_pandas_read_supports_gzip():
with filetext('Alice,1\nBob,2', open=gzip.open,
mode='wt', extension='.csv.gz') as fn:
ds = datashape.dshape('var * {name: string, amount: int}')
csv = CSV(fn)
df = csv_to_dataframe(csv, dshape=ds)
assert isinstance(df, pd.DataFrame)
assert convert(list, df) == [('Alice', 1), ('Bob', 2)]
assert list(df.columns) == ['name', 'amount']
def test_header_mix_str_digits():
ds = datashape.dshape('''var * {"On- or Off- Budget": ?string,
"1990": ?string}''')
with filetext(
'On- or Off- Budget,1990\nOn Budget,-628\nOff budget,"5,962"\n'
) as fn:
csv = CSV(fn, has_header=True)
df = convert(pd.DataFrame, csv)
assert discover(csv).measure == ds.measure
def test_glob():
d = {'accounts1.csv': 'name,when\nAlice,100\nBob,200',
'accounts2.csv': 'name,when\nAlice,300\nBob,400'}
with filetexts(d) as fns:
r = resource('accounts*.csv', has_header=True)
assert convert(list, r) == [('Alice', 100), ('Bob', 200),
('Alice', 300), ('Bob', 400)]
r = resource('*.csv')
assert isinstance(r, chunks(CSV))
def test_csv_append():
with tmpfile('.csv') as fn:
csv = CSV(fn, has_header=False)
data = [('Alice', 100), ('Bob', 200)]
append(csv, data)
assert list(convert(Iterator, csv)) == data
with open(fn) as f:
s = f.read()
assert 'Alice' in s
assert '100' in s
def test_csv_append():
with tmpfile(".csv") as fn:
csv = CSV(fn, has_header=False)
data = [("Alice", 100), ("Bob", 200)]
append(csv, data)
assert list(convert(Iterator, csv)) == data
with open(fn) as f:
s = f.read()
assert "Alice" in s
assert "100" in s
def test_csv_separator_header():
with filetext('a|b|c\n1|2|3\n4|5|6', extension='csv') as fn:
csv = CSV(fn, delimiter='|', has_header=True)
assert convert(list, csv) == [(1, 2, 3), (4, 5, 6)]
def test_empty_dataframe():
with filetext('name,val', extension='csv') as fn:
csv = CSV(fn, has_header=True)
df = convert(pd.DataFrame, csv)
assert isinstance(df, pd.DataFrame)
def test_unused_datetime_columns():
ds = datashape.dshape('var * {val: string, when: datetime}')
with filetext("val,when\na,2000-01-01\nb,2000-02-02") as fn:
csv = CSV(fn, has_header=True)
assert convert(list, csv_to_dataframe(csv, usecols=['val'],
squeeze=True, dshape=ds)) == ['a', 'b']
def test_discover_csv_files_without_header():
with filetext('Alice,2014-01-01\nBob,2014-02-02') as fn:
csv = CSV(fn, has_header=False)
df = convert(pd.DataFrame, csv)
assert len(df) == 2
assert 'Alice' not in list(df.columns)
