def test_het_goldfeldquandt(self):
#TODO: test options missing
#> gq = gqtest(fm, alternative='greater')
#> mkhtest_f(gq, 'het_gq_greater', 'f')
het_gq_greater = dict(statistic=0.5313259064778423,
pvalue=0.9990217851193723,
parameters=(98, 98), distr='f')
#> gq = gqtest(fm, alternative='less')
#> mkhtest_f(gq, 'het_gq_less', 'f')
het_gq_less = dict(statistic=0.5313259064778423,
pvalue=0.000978214880627621,
parameters=(98, 98), distr='f')
#> gq = gqtest(fm, alternative='two.sided')
#> mkhtest_f(gq, 'het_gq_two_sided', 'f')
het_gq_two_sided = dict(statistic=0.5313259064778423,
pvalue=0.001956429761255241,
parameters=(98, 98), distr='f')
#> gq = gqtest(fm, fraction=0.1, alternative='two.sided')
#> mkhtest_f(gq, 'het_gq_two_sided_01', 'f')
het_gq_two_sided_01 = dict(statistic=0.5006976835928314,
pvalue=0.001387126702579789,
parameters=(88, 87), distr='f')
#> gq = gqtest(fm, fraction=0.5, alternative='two.sided')
#> mkhtest_f(gq, 'het_gq_two_sided_05', 'f')
het_gq_two_sided_05 = dict(statistic=0.434815645134117,
pvalue=0.004799321242905568,
parameters=(48, 47), distr='f')
endogg, exogg = self.endog, self.exog
#tests
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=0.5)
compare_t_est(gq, het_gq_greater, decimal=(14, 14))
assert_equal(gq[-1], 'increasing')
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=0.5,
alternative='decreasing')
compare_t_est(gq, het_gq_less, decimal=(14, 14))
assert_equal(gq[-1], 'decreasing')
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=0.5,
alternative='two-sided')
compare_t_est(gq, het_gq_two_sided, decimal=(14, 14))
assert_equal(gq[-1], 'two-sided')
#TODO: forcing the same split as R 202-90-90-1=21
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=90, drop=21,
alternative='two-sided')
compare_t_est(gq, het_gq_two_sided_01, decimal=(14, 14))
assert_equal(gq[-1], 'two-sided')
def test_het_goldfeldquandt(self):
#TODO: test options missing
#> gq = gqtest(fm, alternative='greater')
#> mkhtest_f(gq, 'het_gq_greater', 'f')
het_gq_greater = dict(statistic=0.5313259064778423,
pvalue=0.9990217851193723,
parameters=(98, 98), distr='f')
#> gq = gqtest(fm, alternative='less')
#> mkhtest_f(gq, 'het_gq_less', 'f')
het_gq_less = dict(statistic=0.5313259064778423,
pvalue=0.000978214880627621,
parameters=(98, 98), distr='f')
#> gq = gqtest(fm, alternative='two.sided')
#> mkhtest_f(gq, 'het_gq_two_sided', 'f')
het_gq_two_sided = dict(statistic=0.5313259064778423,
pvalue=0.001956429761255241,
parameters=(98, 98), distr='f')
#> gq = gqtest(fm, fraction=0.1, alternative='two.sided')
#> mkhtest_f(gq, 'het_gq_two_sided_01', 'f')
het_gq_two_sided_01 = dict(statistic=0.5006976835928314,
pvalue=0.001387126702579789,
parameters=(88, 87), distr='f')
#> gq = gqtest(fm, fraction=0.5, alternative='two.sided')
#> mkhtest_f(gq, 'het_gq_two_sided_05', 'f')
het_gq_two_sided_05 = dict(statistic=0.434815645134117,
pvalue=0.004799321242905568,
parameters=(48, 47), distr='f')
endogg, exogg = self.endog, self.exog
#tests
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=0.5)
compare_t_est(gq, het_gq_greater, decimal=(14, 14))
assert_equal(gq[-1], 'increasing')
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=0.5,
alternative='decreasing')
compare_t_est(gq, het_gq_less, decimal=(14, 14))
assert_equal(gq[-1], 'decreasing')
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=0.5,
alternative='two-sided')
compare_t_est(gq, het_gq_two_sided, decimal=(14, 14))
assert_equal(gq[-1], 'two-sided')
#TODO: forcing the same split as R 202-90-90-1=21
gq = smsdia.het_goldfeldquandt(endogg, exogg, split=90, drop=21,
alternative='two-sided')
compare_t_est(gq, het_gq_two_sided_01, decimal=(14, 14))
assert_equal(gq[-1], 'two-sided')
def notyet_atst():
d = macrodata.load().data
realinv = d['realinv']
realgdp = d['realgdp']
realint = d['realint']
endog = realinv
exog = add_constant(np.c_[realgdp, realint],prepend=True)
res_ols1 = OLS(endog, exog).fit()
#growth rates
gs_l_realinv = 400 * np.diff(np.log(d['realinv']))
gs_l_realgdp = 400 * np.diff(np.log(d['realgdp']))
lint = d['realint'][:-1]
tbilrate = d['tbilrate'][:-1]
endogg = gs_l_realinv
exogg = add_constant(np.c_[gs_l_realgdp, lint], prepend=True)
exogg2 = add_constant(np.c_[gs_l_realgdp, tbilrate], prepend=True)
res_ols = OLS(endogg, exogg).fit()
res_ols2 = OLS(endogg, exogg2).fit()
#the following were done accidentally with res_ols1 in R,
#with original Greene data
params = np.array([-272.3986041341653, 0.1779455206941112,
0.2149432424658157])
cov_hac_4 = np.array([1321.569466333051, -0.2318836566017612,
37.01280466875694, -0.2318836566017614, 4.602339488102263e-05,
-0.0104687835998635, 37.012804668757, -0.0104687835998635,
21.16037144168061]).reshape(3,3, order='F')
cov_hac_10 = np.array([2027.356101193361, -0.3507514463299015,
54.81079621448568, -0.350751446329901, 6.953380432635583e-05,
-0.01268990195095196, 54.81079621448564, -0.01268990195095195,
22.92512402151113]).reshape(3,3, order='F')
#goldfeld-quandt
het_gq_greater = dict(statistic=13.20512768685082, df1=99, df2=98,
pvalue=1.246141976112324e-30, distr='f')
het_gq_less = dict(statistic=13.20512768685082, df1=99, df2=98, pvalue=1.)
het_gq_2sided = dict(statistic=13.20512768685082, df1=99, df2=98,
pvalue=1.246141976112324e-30, distr='f')
#goldfeld-quandt, fraction = 0.5
het_gq_greater_2 = dict(statistic=87.1328934692124, df1=48, df2=47,
pvalue=2.154956842194898e-33, distr='f')
gq = smsdia.het_goldfeldquandt(endog, exog, split=0.5)
compare_t_est(gq, het_gq_greater, decimal=(13, 14))
assert_equal(gq[-1], 'increasing')
harvey_collier = dict(stat=2.28042114041313, df=199,
pvalue=0.02364236161988260, distr='t')
#hc = harvtest(fm, order.by=ggdp , data = list())
harvey_collier_2 = dict(stat=0.7516918462158783, df=199,
pvalue=0.4531244858006127, distr='t')
def notyet_atst():
d = macrodata.load(as_pandas=False).data
realinv = d['realinv']
realgdp = d['realgdp']
realint = d['realint']
endog = realinv
exog = add_constant(np.c_[realgdp, realint])
res_ols1 = OLS(endog, exog).fit()
#growth rates
gs_l_realinv = 400 * np.diff(np.log(d['realinv']))
gs_l_realgdp = 400 * np.diff(np.log(d['realgdp']))
lint = d['realint'][:-1]
tbilrate = d['tbilrate'][:-1]
endogg = gs_l_realinv
exogg = add_constant(np.c_[gs_l_realgdp, lint])
exogg2 = add_constant(np.c_[gs_l_realgdp, tbilrate])
res_ols = OLS(endogg, exogg).fit()
res_ols2 = OLS(endogg, exogg2).fit()
#the following were done accidentally with res_ols1 in R,
#with original Greene data
params = np.array([-272.3986041341653, 0.1779455206941112,
0.2149432424658157])
cov_hac_4 = np.array([1321.569466333051, -0.2318836566017612,
37.01280466875694, -0.2318836566017614, 4.602339488102263e-05,
-0.0104687835998635, 37.012804668757, -0.0104687835998635,
21.16037144168061]).reshape(3,3, order='F')
cov_hac_10 = np.array([2027.356101193361, -0.3507514463299015,
54.81079621448568, -0.350751446329901, 6.953380432635583e-05,
-0.01268990195095196, 54.81079621448564, -0.01268990195095195,
22.92512402151113]).reshape(3,3, order='F')
#goldfeld-quandt
het_gq_greater = dict(statistic=13.20512768685082, df1=99, df2=98,
pvalue=1.246141976112324e-30, distr='f')
het_gq_less = dict(statistic=13.20512768685082, df1=99, df2=98, pvalue=1.)
het_gq_2sided = dict(statistic=13.20512768685082, df1=99, df2=98,
pvalue=1.246141976112324e-30, distr='f')
#goldfeld-quandt, fraction = 0.5
het_gq_greater_2 = dict(statistic=87.1328934692124, df1=48, df2=47,
pvalue=2.154956842194898e-33, distr='f')
gq = smsdia.het_goldfeldquandt(endog, exog, split=0.5)
compare_t_est(gq, het_gq_greater, decimal=(13, 14))
assert_equal(gq[-1], 'increasing')
harvey_collier = dict(stat=2.28042114041313, df=199,
pvalue=0.02364236161988260, distr='t')
#hc = harvtest(fm, order.by=ggdp , data = list())
harvey_collier_2 = dict(stat=0.7516918462158783, df=199,
pvalue=0.4531244858006127, distr='t')
def print_goldsfeltquandt(self):
linear_model = sm.OLS(self.y_train, self.X_train).fit()
print(
diagnostic.het_goldfeldquandt(linear_model.resid,
linear_model.model.exog))
plt.xlabel('NINO3.4 ($^{\circ}$C)')
plt.ylabel('Residuals (mm day$^{-1}$)')
plt.title('Scatter Plot: Residuals vs. NINO3.4')
# It's a bit hard to tell from this plot. Let's try the Goldfeld-Quandt test. To use this test we need to import a new python package: `statsmodels`.
# In[7]:
# test for homoscedasticity
from statsmodels.stats.diagnostic import het_goldfeldquandt
nino34ex = np.expand_dims(nino34,
axis=1) # we need to reshape nino34 to use this test
# the output of the het_goldfeldquandt test is the F-statistic and the p-value.
GQtest = het_goldfeldquandt(precip_djfm, nino34ex, alternative='increasing')
print(GQtest[1])
# Thus, we can accept the null hypothesis of homoscedasticity given the large p-value.
# ## 2. No Autocorrelation of Residuals (Durbin-Watson test)
#
# The condition that the residuals should be independent means that one residual should not be able to predict the next. But, if the residuals contain a trend or periodicity, then the residuals are not truly independent and are **autocorrelated** (see [Figure 15](autocorr)).
#
# ```{figure} autocorr.png
# ---
# scale: 50%
# name: autocorr
# ---
# Scatter plots of the residuals versus $x$ for residuals the have (left) no autocorrelation and (right) autocorrelation.
# ```
def smf_het_goldfeldquandt(self):
f_stat, p_val, inc_dec = het_goldfeldquandt(self.model1.resid,
self.model1.model.exog)
print(
f'For smf model 1 het goldfeldquandt test, the f stat is {f_stat} and the p value is {p_val}'
)
diag.het_goldfeldquandt(modelMR2.resid, modelMR2.model.exog) diag.het_breuschpagan(modelMR4.resid, modelMR4.model.exog) diag.het_white(modelMR2.resid, modelMR2.model.exog, retres = False) diag.(modelMR2.resid, modelMR2.model.exog) diag.acorr_ljungbox(modelMR2.resid)
