def testCompleteSeriesFunctionalInput(self):
checkpoints.flush()
s1 = pd.Series([1, 2, 3]).safe_map(lambda v: pd.Series([1,2]))
s2 = pd.Series([1, 2, 3]).map(lambda v: pd.Series([1,2]))
self.assertTrue(isinstance(s1[0], pd.Series) and isinstance(s2[0], pd.Series))
self.assertTrue(np.array_equal(s1[0].values, s2[0].values))
self.assertTrue(np.array_equal(s1[1].values, s2[1].values))
def testCompleteDictFunctionalInput(self):
checkpoints.flush()
d1 = pd.Series([1, 2, 3]).safe_map(lambda v: pd.Series({'a': 1, 'b': 2}))
d2 = pd.Series([1, 2, 3]).map(lambda v: pd.Series({'a': 1, 'b': 2}))
self.assertTrue(isinstance(d1[0], pd.Series) and isinstance(d2[0], pd.Series))
self.assertTrue(np.array_equal(d1[0].values, d1[0].values))
self.assertTrue(np.array_equal(d2[1].values, d2[1].values))
def testFlush(self):
checkpoints._results = ['foo', 'bar']
checkpoints._index = pd.Index([1,2])
checkpoints.flush()
self.assertEqual(checkpoints._results, [])
self.assertEqual(checkpoints._index, None)
pass
def testFunctionWithKwargs(self):
checkpoints.flush()
def f(srs, **kwargs):
return 1 + sum(kwargs.values())
r = pd.DataFrame({'a': [1, 2], 'b': [3, 4]}).safe_apply(f, somearg=5)
self.assertTrue(np.array_equal(r.values, [6, 6]))
def testPartialScalarFunctionalInput(self):
checkpoints.flush()
df = pd.DataFrame(np.random.random(100).reshape((20, 5)))
df[2][4] = 0
def breaker(srs):
if 0 in srs.values:
raise IOError
else:
return srs.sum()
import warnings
warnings.filterwarnings('ignore')
with self.assertRaises(IOError):
df.safe_apply(breaker, axis='columns')
self.assertEquals(len(checkpoints._results), 4)
self.assertIsNot(checkpoints._index, None)
def testPartialScalarFunctionalInput(self):
checkpoints.flush()
srs = pd.Series(np.random.random(100))
srs[50] = 0
def breaker(val):
if val == 0:
raise IOError
else:
return val
import warnings
warnings.filterwarnings('ignore')
with self.assertRaises(IOError):
srs.safe_map(breaker)
self.assertEquals(len(checkpoints._results), 50)
self.assertIsNot(checkpoints._index, None)
srs[50] = 1
result = srs.safe_map(breaker)
self.assertIsNot(result, None)
self.assertIsInstance(result, pd.Series)
self.assertEqual(len(result), 100)
def testCompleteListFunctionalInput(self):
checkpoints.flush()
self.assertTrue(
np.array_equal(pd.Series([1,2,3]).safe_map(lambda v: [1,2]),
pd.Series([1, 2, 3]).map(lambda v: [1, 2]))
)
def testCompleteScalarFunctionalInput(self):
checkpoints.flush()
srs = pd.Series(np.random.random(100))
m1 = np.average(srs.map(lambda val: val - 0.5))
m2 = np.average(srs.safe_map(lambda val: val - 0.5))
self.assertAlmostEqual(m1, m2)
def testResultsEmpty(self):
checkpoints.flush()
self.assertEqual(checkpoints.results, None)
def testCompleteDictFunctionalInputColumnar(self):
checkpoints.flush()
d1 = pd.DataFrame({'a': [1, 2], 'b': [3, 4]}).safe_apply(lambda v: pd.Series({'a': 1, 'b': 2}), axis=1)
d2 = pd.DataFrame({'a': [1, 2], 'b': [3, 4]}).apply(lambda v: pd.Series({'a': 1, 'b': 2}), axis=1)
self.assertTrue(d1.equals(d2))
def testCompleteSeriesFunctionalInputColumnar(self):
checkpoints.flush()
s1 = pd.DataFrame({'a': [1, 2], 'b': [3, 4]}).safe_apply(lambda v: pd.Series(['A','B']), axis=1)
s2 = pd.DataFrame({'a': [1, 2], 'b': [3, 4]}).apply(lambda v: pd.Series(['A','B']), axis=1)
self.assertTrue(s1.equals(s2))
def testCompleteListFunctionalInput(self):
checkpoints.flush()
self.assertTrue(
np.array_equal(pd.DataFrame({'a': [1, 2], 'b': [3, 4]}).safe_apply(lambda v: [1, 2]),
pd.DataFrame({'a': [1, 2], 'b': [3, 4]}).safe_apply(lambda v: [1, 2]))
)
def testCompleteScalarFunctionalInput(self):
checkpoints.flush()
df = pd.DataFrame(np.random.random(100).reshape((20, 5)))
m1 = np.average(df.apply(sum))
m2 = np.average(df.safe_apply(sum))
self.assertAlmostEqual(m1, m2)