def test_limited_redundancy():
data = generate_data(0,
datatype="numpy",
const=False,
other_effects=1,
ntk=(25, 200, 5))
for i in range(0, data.x.shape[1], 7):
data.x[1, i, :] = data.x[0, i, :]
mod = FamaMacBeth(data.y, data.x)
res = mod.fit()
assert np.any(np.isnan(res.all_params))
def test_block_size_warnings():
y = np.arange(12.0)[:, None]
x = np.ones((12, 3))
x[:, 1] = np.arange(12.0)
x[:, 2] = np.arange(12.0)**2
idx = pd.MultiIndex.from_product([["a", "b", "c"],
pd.date_range("2000-1-1", periods=4)])
y = pd.DataFrame(y, index=idx, columns=["y"])
x = pd.DataFrame(x, index=idx, columns=["x1", "x2", "x3"])
with pytest.warns(MissingValueWarning):
FamaMacBeth(y.iloc[:11], x.iloc[:11])
with pytest.warns(InferenceUnavailableWarning):
FamaMacBeth(y.iloc[::4], x.iloc[::4])
def test_block_size_warnings():
y = np.arange(12.0)[:, None]
x = np.ones((12, 3))
x[:, 1] = np.arange(12.0)
x[:, 2] = np.arange(12.0)**2
idx = pd.MultiIndex.from_product([['a', 'b', 'c'],
pd.date_range('2000-1-1', periods=4)])
y = pd.DataFrame(y, index=idx, columns=['y'])
x = pd.DataFrame(x, index=idx, columns=['x1', 'x2', 'x3'])
with pytest.warns(MissingValueWarning):
FamaMacBeth(y.iloc[:11], x.iloc[:11])
with pytest.warns(InferenceUnavailableWarning):
FamaMacBeth(y.iloc[::4], x.iloc[::4])
def test_fama_macbeth(data):
res = FamaMacBeth(data.y, data.x).fit(debiased=True)
y = PanelData(data.y)
x = PanelData(data.x)
missing = y.isnull | x.isnull
y.drop(missing)
x.drop(missing)
y = y.dataframe
x = x.dataframe
times = y.index.levels[1]
params = []
for t in times:
_y = y.xs(t, level=1)
_x = x.xs(t, level=1)
if _x.shape[0] < _x.shape[1]:
continue
_x = _x.loc[_y.index]
params.append(np.linalg.lstsq(_x.values, _y.values)[0])
params = np.array(params).squeeze()
all_params = params
params = params.mean(0)
assert_allclose(params.squeeze(), res.params)
e_params = all_params - params[None, :]
ntime = e_params.shape[0]
cov = e_params.T @ e_params / ntime / (ntime - 1)
assert_allclose(cov, res.cov.values)
d = dir(res)
for key in d:
if not key.startswith('_'):
val = getattr(res, key)
if callable(val):
val()
def test_fama_macbeth(data):
res = FamaMacBeth(data.y, data.x).fit(debiased=True)
y = PanelData(data.y)
x = PanelData(data.x)
missing = y.isnull | x.isnull
y.drop(missing)
x.drop(missing)
y = y.dataframe
x = x.dataframe
times = y.index.levels[1]
params = []
for t in times:
_y = y.xs(t, level=1)
_x = x.xs(t, level=1)
if _x.shape[0] < _x.shape[1]:
continue
_x = _x.loc[_y.index]
params.append(lstsq(_x.values, _y.values, rcond=None)[0])
params = np.array(params).squeeze()
all_params = params
params = params.mean(0)
assert_allclose(params.squeeze(), res.params)
assert_allclose(all_params, res.all_params.dropna(how="all"))
e_params = all_params - params[None, :]
ntime = e_params.shape[0]
cov = e_params.T @ e_params / ntime / (ntime - 1)
assert_allclose(cov, np.asarray(res.cov))
access_attributes(res)
def test_fama_macbeth_kernel_smoke(data):
FamaMacBeth(data.y, data.x).fit(cov_type="kernel")
FamaMacBeth(data.y, data.x).fit(cov_type="kernel", kernel="bartlett")
FamaMacBeth(data.y, data.x).fit(cov_type="kernel", kernel="newey-west")
FamaMacBeth(data.y, data.x).fit(cov_type="kernel", kernel="parzen")
FamaMacBeth(data.y, data.x).fit(cov_type="kernel", kernel="qs")
FamaMacBeth(data.y, data.x).fit(cov_type="kernel", bandwidth=3)
res = FamaMacBeth(data.y, data.x).fit(cov_type="kernel", kernel="andrews")
access_attributes(res)
def test_fama_macbeth_kernel_smoke(data):
FamaMacBeth(data.y, data.x).fit(cov_type='kernel')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='bartlett')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='newey-west')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='parzen')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='qs')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', bandwidth=3)
res = FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='andrews')
access_attributes(res)
def test_fitted_effects_residuals(data):
mod = FamaMacBeth(data.y, data.x)
res = mod.fit()
expected = mod.exog.values2d @ res.params.values
expected = pd.DataFrame(expected,
index=mod.exog.index,
columns=['fitted_values'])
assert_allclose(res.fitted_values, expected)
assert_frame_similar(res.fitted_values, expected)
expected.iloc[:, 0] = mod.dependent.values2d - expected.values
expected.columns = ['idiosyncratic']
assert_allclose(res.idiosyncratic, expected)
assert_frame_similar(res.idiosyncratic, expected)
expected.iloc[:, 0] = np.nan
expected.columns = ['estimated_effects']
assert_allclose(res.estimated_effects, expected)
assert_frame_similar(res.estimated_effects, expected)
def test_block_size_error():
y = np.arange(12.0)[:, None]
x = np.ones((12, 2))
x[1::4, 1] = 2
x[2::4, 1] = 3
idx = pd.MultiIndex.from_product([["a", "b", "c"],
pd.date_range("2000-1-1", periods=4)])
y = pd.DataFrame(y, index=idx, columns=["y"])
x = pd.DataFrame(x, index=idx, columns=["x1", "x2"])
with pytest.raises(ValueError):
FamaMacBeth(y, x)
def test_block_size_error():
y = np.arange(12.0)[:, None]
x = np.ones((12, 2))
x[1::4, 1] = 2
x[2::4, 1] = 3
idx = pd.MultiIndex.from_product([['a', 'b', 'c'],
pd.date_range('2000-1-1', periods=4)])
y = pd.DataFrame(y, index=idx, columns=['y'])
x = pd.DataFrame(x, index=idx, columns=['x1', 'x2'])
with pytest.raises(ValueError):
FamaMacBeth(y, x)
def test_fama_macbeth_kernel_smoke(data):
FamaMacBeth(data.y, data.x).fit(cov_type='kernel')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='bartlett')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='newey-west')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='parzen')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='qs')
FamaMacBeth(data.y, data.x).fit(cov_type='kernel', bandwidth=3)
res = FamaMacBeth(data.y, data.x).fit(cov_type='kernel', kernel='andrews')
d = dir(res)
for key in d:
if not key.startswith('_'):
val = getattr(res, key)
if callable(val):
val()
def test_unknown_cov_type(data):
with pytest.raises(ValueError):
FamaMacBeth(data.y, data.x).fit(cov_type='unknown')
