def test_groupby(self):
df20 = dummy_streaming_dataframe(20).to_dataframe()
df20["key"] = df20["cint"].apply(lambda i: i % 3 == 0)
sdf20 = StreamingDataFrame.read_df(df20, chunksize=5)
gr = sdf20.groupby("key", lambda gr: gr.sum())
gr2 = df20.groupby("key").sum()
self.assertEqualDataFrame(gr, gr2)
self.assertRaise(lambda: sdf20.groupby(
"key", in_memory=False), NotImplementedError)
# Do not replace lambda c:sum(c) by sum or...
# pandas.core.base.SpecificationError: Function names must be unique, found multiple named sum
gr2 = df20.groupby("key").agg([numpy.sum, lambda c:sum(c)])
gr = sdf20.groupby("key", lambda gr: gr.agg(
[numpy.sum, lambda c:sum(c)]))
self.assertEqualDataFrame(gr, gr2)
gr = sdf20.groupby("key", lambda gr: gr.count())
gr2 = df20.groupby("key").count()
self.assertEqualDataFrame(gr, gr2)
df = pandas.DataFrame(dict(A=[3, 4, 3], B=[5, 6, 7]))
sdf = StreamingDataFrame.read_df(df)
gr = sdf.groupby("A")
gr2 = df.groupby("A").sum()
self.assertEqualDataFrame(gr, gr2)
def test_read_csv(self):
temp = get_temp_folder(__file__, "temp_read_csv")
df = pandas.DataFrame(data=dict(a=[5, 6], b=["er", "r"]))
name = os.path.join(temp, "df.csv")
name2 = os.path.join(temp, "df2.csv")
name3 = os.path.join(temp, "df3.csv")
df.to_csv(name, index=False)
df.to_csv(name2, index=True)
sdf = StreamingDataFrame.read_csv(name)
text = sdf.to_csv(index=False)
self.assertRaise(
lambda: StreamingDataFrame.read_csv(
name2, index_col=0, chunksize=None),
ValueError)
self.assertRaise(
lambda: StreamingDataFrame.read_csv(
name2, index_col=0, iterator=False),
ValueError)
sdf2 = StreamingDataFrame.read_csv(name2, index_col=0)
text2 = sdf2.to_csv(index=True)
sdf2.to_csv(name3, index=True)
with open(name, "r", encoding='utf-8') as f:
exp = f.read()
with open(name2, "r", encoding='utf-8') as f:
exp2 = f.read()
with open(name3, "r", encoding='utf-8') as f:
text3 = f.read()
self.assertEqual(text.replace('\r', ''), exp)
sdf2 = StreamingDataFrame.read_df(df)
self.assertEqualDataFrame(sdf.to_dataframe(), sdf2.to_dataframe())
self.assertEqual(text2.replace('\r', ''), exp2)
self.assertEqual(text3.replace('\r', '').replace('\n\n', '\n'),
exp2.replace('\r', ''))
def test_set_item_function(self):
df = pandas.DataFrame(data=dict(a=[4.5], b=[6], c=[7]))
self.assertRaise(lambda: StreamingDataFrame(df), TypeError)
sdf = StreamingDataFrame.read_df(df)
sdf['bb'] = sdf['b'].apply(lambda x: x + 11)
df = sdf.to_df()
ddf = ddf = pandas.DataFrame(
data=dict(a=[4.5], b=[6], c=[7], bb=[17]))
self.assertEqualDataFrame(df, ddf)
def test_add_column(self):
df = pandas.DataFrame(data=dict(X=[4.5, 6, 7], Y=["a", "b", "c"]))
sdf = StreamingDataFrame.read_df(df)
sdf2 = sdf.add_column("d", lambda row: int(1))
df2 = sdf2.to_dataframe()
df["d"] = 1
self.assertEqualDataFrame(df, df2)
sdf3 = StreamingDataFrame.read_df(df)
sdf4 = sdf3.add_column("dd", 2)
df4 = sdf4.to_dataframe()
df["dd"] = 2
self.assertEqualDataFrame(df, df4)
sdfA = StreamingDataFrame.read_df(df)
sdfB = sdfA.add_column("dd12", lambda row: row["dd"] + 10)
dfB = sdfB.to_dataframe()
df["dd12"] = 12
self.assertEqualDataFrame(df, dfB)
def test_groupby_streaming(self):
df20 = dummy_streaming_dataframe(20).to_dataframe()
df20["key"] = df20["cint"].apply(lambda i: i % 3 == 0)
sdf20 = StreamingDataFrame.read_df(df20, chunksize=5)
sgr = sdf20.groupby_streaming(
"key", lambda gr: gr.sum(), strategy='streaming', as_index=False)
gr2 = df20.groupby("key", as_index=False).sum()
grs = list(sgr)
gr = pandas.concat(grs).groupby("key", as_index=False).sum()
self.assertEqualDataFrame(gr, gr2)
def test_merge_2(self):
df = pandas.DataFrame(data=dict(X=[4.5, 6, 7], Y=["a", "b", "c"]))
df2 = pandas.concat([df, df])
sdf = StreamingDataFrame.read_df(df)
sdf2 = sdf.concat(sdf, axis=0)
self.assertEqualDataFrame(df2, sdf2.to_dataframe())
self.assertEqualDataFrame(df2, sdf2.to_dataframe())
m = pandas.DataFrame(dict(Y=["a", "b"], Z=[10, 20]))
jm = df2.merge(m, left_on="Y", right_on="Y", how="outer")
sjm = sdf2.merge(m, left_on="Y", right_on="Y", how="outer")
self.assertEqualDataFrame(jm.sort_values(["X", "Y"]).reset_index(drop=True),
sjm.to_dataframe().sort_values(["X", "Y"]).reset_index(drop=True))
def test_groupby_cum_asindex(self):
df20 = dummy_streaming_dataframe(20).to_dataframe()
df20["key"] = df20["cint"].apply(lambda i: i % 3 == 0)
sdf20 = StreamingDataFrame.read_df(df20, chunksize=5)
sgr = sdf20.groupby_streaming(
"key", lambda gr: gr.sum(), strategy='cum', as_index=True)
gr2 = df20.groupby("key", as_index=True).sum()
lastgr = None
for gr in sgr:
self.assertEqual(list(gr.columns), list(gr2.columns))
lastgr = gr
self.assertEqualDataFrame(lastgr, gr2)
def test_sort_values_reverse(self):
temp = get_temp_folder(__file__, "temp_sort_values_reverse")
name = os.path.join(temp, "_data_")
df = pandas.DataFrame([dict(a=1, b="eé", c=5.6, ind="a1", ai=1),
dict(a=5, b="f", c=5.7, ind="a2", ai=2),
dict(a=4, b="g", ind="a3", ai=3),
dict(a=8, b="h", c=5.9, ai=4),
dict(a=16, b="i", c=6.2, ind="a5", ai=5)])
sdf = StreamingDataFrame.read_df(df, chunksize=2)
sorted_df = df.sort_values(by="a", ascending=False)
res = sdf.sort_values(by="a", temp_file=name, ascending=False)
res_df = res.to_df()
self.assertEqualDataFrame(sorted_df, res_df)
def test_train_test_split_streaming_tiny(self):
df = pandas.DataFrame(data=dict(X=[4.5, 6, 7], Y=["a", "b", "c"]))
sdf2 = StreamingDataFrame.read_df(pandas.concat([df, df]))
sdftr, sdfte = sdf2.train_test_split(test_size=0.5)
df1 = sdfte.head()
df2 = sdfte.head()
if df1 is not None or df2 is not None:
self.assertEqualDataFrame(df1, df2)
df1 = sdftr.head()
df2 = sdftr.head()
if df1 is not None or df2 is not None:
self.assertEqualDataFrame(df1, df2)
sdf = StreamingDataFrame.read_df(df)
sdf2 = sdf.concat(sdf, axis=0)
sdftr, sdfte = sdf2.train_test_split(test_size=0.5)
df1 = sdfte.head()
df2 = sdfte.head()
if df1 is not None or df2 is not None:
self.assertEqualDataFrame(df1, df2)
df1 = sdftr.head()
df2 = sdftr.head()
if df1 is not None or df2 is not None:
self.assertEqualDataFrame(df1, df2)
def test_fillna(self):
df = pandas.DataFrame(
data=dict(X=[4.5, numpy.nan, 7], Y=["a", "b", numpy.nan]))
sdf = StreamingDataFrame.read_df(df)
df2 = pandas.DataFrame(
data=dict(X=[4.5, 10.0, 7], Y=["a", "b", "NAN"]))
na = sdf.fillna(value=dict(X=10.0, Y="NAN"))
ndf = na.to_df()
self.assertEqual(ndf, df2)
df3 = pandas.DataFrame(
data=dict(X=[4.5, 10.0, 7], Y=["a", "b", numpy.nan]))
na = sdf.fillna(value=dict(X=10.0))
ndf = na.to_df()
self.assertEqual(ndf, df3)
def test_describe(self):
x = numpy.arange(100001).astype(numpy.float64) / 100000 - 0.5
y = numpy.arange(100001).astype(numpy.int64)
z = numpy.array([chr(65 + j % 45) for j in y])
df = pandas.DataFrame(data=dict(X=x, Y=y, Z=z))
sdf = StreamingDataFrame.read_df(df)
desc = sdf.describe()
self.assertEqual(['X', 'Y'], list(desc.columns))
self.assertEqual(desc.loc['min', :].tolist(), [-0.5, 0])
self.assertEqual(desc.loc['max', :].tolist(), [0.5, 100000])
self.assertEqualArray(desc.loc['mean', :], numpy.array([0, 50000]))
self.assertEqualArray(desc.loc['25%', :], numpy.array([-0.25, 25000]))
self.assertEqualArray(desc.loc['50%', :], numpy.array([0.0, 50000]))
self.assertEqualArray(desc.loc['75%', :], numpy.array([0.25, 75000]))
self.assertEqualArray(desc.loc['std', :], numpy.array(
[2.886795e-01, 28867.946472]), decimal=4)
def test_set_item(self):
df = pandas.DataFrame(data=dict(a=[4.5], b=[6], c=[7]))
self.assertRaise(lambda: StreamingDataFrame(df), TypeError)
sdf = StreamingDataFrame.read_df(df)
def f():
sdf[['a']] = 10
self.assertRaise(f, ValueError)
def g():
sdf['a'] = [10]
self.assertRaise(g, NotImplementedError)
sdf['aa'] = 10
df = sdf.to_df()
ddf = pandas.DataFrame(data=dict(a=[4.5], b=[6], c=[7], aa=[10]))
self.assertEqualDataFrame(df, ddf)
sdf['bb'] = sdf['b'] + 10
df = sdf.to_df()
ddf = ddf = pandas.DataFrame(
data=dict(a=[4.5], b=[6], c=[7], aa=[10], bb=[16]))
self.assertEqualDataFrame(df, ddf)
class MyStdout(object):
def __init__(self, term=sys.stdout):
self.term = term
def write(self, text):
text = colorize(text)
self.term.write(text)
def flush(self):
pass
sys.stdout = MyStdout()
from pandas_streaming.df import StreamingDataFrame
sdf = StreamingDataFrame.read_df(data)
for data in sdf:
print(data)
print("")
active = df.status.str.count("active").sum()
inactive = df.status.str.count("inactive").sum()
invalid = df.status.str.count("invalid").sum()
print("\033[0;32mNumber of active SatNOGS: \033[0m", active)
print("\033[0;31mNumber of inactive SatNOGS: \033[0m", inactive)
print("\033[0;33mNumber of invalid SatNOGS: \033[0m", invalid)
for remaining in range(300, 0, -1):
sys.stdout.write("\r")
sys.stdout.write("Refreshing in {:2d} seconds...".format(remaining))
sys.stdout.flush()