def test_dataframe_equality_with_multi_sort(self):
shuffidx = numpy.arange(size)
shuffle(shuffidx)
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame({ K : self.cols[K][shuffidx] for K in self.cols.keys()})
self.assertTrue(checkeq(df1,df2,sort_columns=["strB","chrA"]))
self.assertTrue(checkeq(df2,df1,sort_columns=["strB","chrA"]))
def test_dataframe_inequality_wrong_rows(self):
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame(
{K: self.cols[K][:size - 1000]
for K in self.cols.keys()})
self.assertFalse(checkeq(df1, df2))
self.assertFalse(checkeq(df2, df1))
def test_dataframe_equality_without_sort(self):
shuffidx = numpy.arange(size)
shuffle(shuffidx)
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame({ K : self.cols[K][shuffidx] for K in self.cols.keys()})
self.assertFalse(checkeq(df1,df2))
self.assertFalse(checkeq(df2,df1))
def test_dataframe_inequality_wrong_columns(self):
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame({
K: copy.deepcopy(self.cols[K])
for K in sorted(self.cols.keys())[:-2]
})
self.assertFalse(checkeq(df1, df2))
self.assertFalse(checkeq(df2, df1))
def test_dataframe_equality_with_multi_sort(self):
shuffidx = numpy.arange(size)
shuffle(shuffidx)
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame(
{K: self.cols[K][shuffidx]
for K in self.cols.keys()})
self.assertTrue(checkeq(df1, df2, sort_columns=["strB", "chrA"]))
self.assertTrue(checkeq(df2, df1, sort_columns=["strB", "chrA"]))
def test_dataframe_equality_without_sort(self):
shuffidx = numpy.arange(size)
shuffle(shuffidx)
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame(
{K: self.cols[K][shuffidx]
for K in self.cols.keys()})
self.assertFalse(checkeq(df1, df2))
self.assertFalse(checkeq(df2, df1))
def test_dataframe_inequality_above_tol(self):
tol = 10**-8
noiseA = tol * 10**2 * numpy.random.randn(size)
noiseB = tol * 10**2 * numpy.random.randn(size)
df1 = pd.DataFrame(self.cols)
df2 = copy.deepcopy(df1)
df2["floatA"] += noiseA
df2["floatB"] -= noiseB
self.assertFalse(checkeq(df1, df2, tol=tol))
self.assertFalse(checkeq(df2, df1, tol=tol))
def test_dataframe_inequality_above_tol(self):
tol = 10**-8
noiseA = tol*10**2 * numpy.random.randn(size)
noiseB = tol*10**2 * numpy.random.randn(size)
df1 = pd.DataFrame(self.cols)
df2 = copy.deepcopy(df1)
df2["floatA"] += noiseA
df2["floatB"] -= noiseB
self.assertFalse(checkeq(df1,df2,tol=tol))
self.assertFalse(checkeq(df2,df1,tol=tol))
def check_dataframe_equality_same(self,special_val):
"""Helper function for checking equality when various special values
are in the same location in both dataframes
Parameters
----------
special_val : numpy.nan, numpy.inf, -numpy.inf, or None
Value that is ignored by :py:meth:`plastid.bin.test_dataframe_equality` if it occurs in the same cells of both dataframes
"""
idx = numpy.random.randint(0,high=size,size=500)
tmpcols = copy.deepcopy(self.cols)
tmpcols["floatA"][idx] = special_val
df1 = pd.DataFrame(tmpcols)
df2 = pd.DataFrame(copy.deepcopy(tmpcols))
self.assertTrue(checkeq(df1,df2))
self.assertTrue(checkeq(df2,df1))
def check_dataframe_equality_same(self, special_val):
"""Helper function for checking equality when various special values
are in the same location in both dataframes
Parameters
----------
special_val : numpy.nan, numpy.inf, -numpy.inf, or None
Value that is ignored by :py:meth:`plastid.bin.test_dataframe_equality` if it occurs in the same cells of both dataframes
"""
idx = numpy.random.randint(0, high=size, size=500)
tmpcols = copy.deepcopy(self.cols)
tmpcols["floatA"][idx] = special_val
df1 = pd.DataFrame(tmpcols)
df2 = pd.DataFrame(copy.deepcopy(tmpcols))
self.assertTrue(checkeq(df1, df2))
self.assertTrue(checkeq(df2, df1))
def check_dataframe_inequality_different(self,special_val):
"""Helper function for checking inequality when various special values
appear in the different locations in both dataframes
Parameters
----------
special_val : numpy.nan, numpy.inf, -numpy.inf, or None
Value that is ignored by :py:meth:`plastid.bin.test_dataframe_equality` if it occurs in the same cells of both dataframes
"""
idxA = numpy.random.randint(0,high=size,size=500)
idxB = numpy.random.randint(0,high=size,size=500)
colsA = copy.deepcopy(self.cols)
colsA["floatA"][idxA] = special_val
colsB = copy.deepcopy(self.cols)
colsB["floatA"][idxB] = special_val
df1 = pd.DataFrame(colsA)
df2 = pd.DataFrame(colsB)
self.assertFalse(checkeq(df1,df2))
self.assertFalse(checkeq(df2,df1))
def check_dataframe_inequality_different(self, special_val):
"""Helper function for checking inequality when various special values
appear in the different locations in both dataframes
Parameters
----------
special_val : numpy.nan, numpy.inf, -numpy.inf, or None
Value that is ignored by :py:meth:`plastid.bin.test_dataframe_equality` if it occurs in the same cells of both dataframes
"""
idxA = numpy.random.randint(0, high=size, size=500)
idxB = numpy.random.randint(0, high=size, size=500)
colsA = copy.deepcopy(self.cols)
colsA["floatA"][idxA] = special_val
colsB = copy.deepcopy(self.cols)
colsB["floatA"][idxB] = special_val
df1 = pd.DataFrame(colsA)
df2 = pd.DataFrame(colsB)
self.assertFalse(checkeq(df1, df2))
self.assertFalse(checkeq(df2, df1))
def test_dataframe_inequality_wrong_rows(self):
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame({K : self.cols[K][:size-1000] for K in self.cols.keys()})
self.assertFalse(checkeq(df1,df2))
self.assertFalse(checkeq(df2,df1))
def test_dataframe_inequality_wrong_columns(self):
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame({ K : copy.deepcopy(self.cols[K]) for K in sorted(self.cols.keys())[:-2] })
self.assertFalse(checkeq(df1,df2))
self.assertFalse(checkeq(df2,df1))
def test_dataframe_quality_when_identical(self):
df1 = pd.DataFrame(self.cols)
self.assertTrue(checkeq(df1, df1))
def test_dataframe_equality_no_sort(self):
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame(copy.deepcopy(self.cols))
self.assertTrue(checkeq(df1,df2))
self.assertTrue(checkeq(df2,df1))
def test_dataframe_quality_when_identical(self):
df1 = pd.DataFrame(self.cols)
self.assertTrue(checkeq(df1,df1))
def test_dataframe_equality_no_sort(self):
df1 = pd.DataFrame(self.cols)
df2 = pd.DataFrame(copy.deepcopy(self.cols))
self.assertTrue(checkeq(df1, df2))
self.assertTrue(checkeq(df2, df1))
