Merge pull request #5805 from jbrockmendel/E4

REF/CLN: collect imports at top of file, de-duplicate imports

setup.cfg

@@ -65,8 +65,6 @@ select=
     # E228: missing whitespace around modulo operator
     E211,
     # E211: whitespace before '['
-    E401,
-    # E401: multiple imports on one line
     W601,
     # W601: .has_key() is deprecated, use 'in'
     W391,
@@ -159,6 +157,10 @@ select=
     # E274: tab before keyword
     # E275: missing whitespace after keyword
 
+    E4,
+    # E401: multiple imports on one line
+    # E402: module level import not at top of file
+
     E9,
     # E901: SyntaxError or IndentationError
     # E902: IOError

statsmodels/discrete/tests/test_constrained.py

@@ -25,12 +25,13 @@
 from statsmodels.tools.tools import add_constant
 from statsmodels import datasets
 
+from .results import results_poisson_constrained as results
+from .results import results_glm_logit_constrained as reslogit
+
 
 spector_data = datasets.spector.load(as_pandas=False)
 spector_data.exog = add_constant(spector_data.exog, prepend=False)
 
-from .results import results_poisson_constrained as results
-from .results import results_glm_logit_constrained as reslogit
 
 DEBUG = False
 

statsmodels/examples/es_misc_poisson2.py

@@ -5,6 +5,8 @@
 import statsmodels.api as sm
 from statsmodels.miscmodels.count import (PoissonGMLE, PoissonOffsetGMLE,
                                           PoissonZiGMLE)
+from statsmodels.discrete.discrete_model import Poisson
+
 
 DEC = 3
 
@@ -23,7 +25,6 @@ class Dummy(object):
 data_endog = np.random.poisson(np.exp(xbeta))
 
 #estimate discretemod.Poisson as benchmark
-from statsmodels.discrete.discrete_model import Poisson
 res_discrete = Poisson(data_endog, data_exog).fit()
 
 mod_glm = sm.GLM(data_endog, data_exog, family=sm.families.Poisson())

statsmodels/examples/ex_generic_mle.py

@@ -6,6 +6,7 @@
 from scipy import stats
 import statsmodels.api as sm
 from statsmodels.base.model import GenericLikelihoodModel
+from statsmodels.tools.numdiff import approx_fprime, approx_hess
 
 
 data = sm.datasets.spector.load(as_pandas=False)
@@ -119,7 +120,6 @@ def loglikeobs(self, params):
 res_bfgs = mod_norm2.fit(start_params=start_params, method="bfgs", fprime=None,
                          maxiter=500, retall=0)
 
-from statsmodels.tools.numdiff import approx_fprime, approx_hess
 hb=-approx_hess(res_norm3.params, mod_norm2.loglike, epsilon=-1e-4)
 hf=-approx_hess(res_norm3.params, mod_norm2.loglike, epsilon=1e-4)
 hh = (hf+hb)/2.

statsmodels/examples/ex_generic_mle_t.py

@@ -12,6 +12,7 @@
 from scipy import special
 import statsmodels.api as sm
 from statsmodels.base.model import GenericLikelihoodModel
+from statsmodels.tools.numdiff import approx_hess
 
 #redefine some shortcuts
 np_log = np.log
@@ -92,7 +93,6 @@ def nloglikeobs(self, params):
 print(resp.params)
 print(resp.bse)
 
-from statsmodels.tools.numdiff import approx_hess
 
 hb=-approx_hess(modp.start_value, modp.loglike, epsilon=-1e-4)
 tmp = modp.loglike(modp.start_value)

statsmodels/examples/ex_generic_mle_tdist.py

@@ -13,6 +13,14 @@
 from scipy import stats, special, optimize
 import statsmodels.api as sm
 from statsmodels.base.model import GenericLikelihoodModel
+from statsmodels.tools.numdiff import approx_hess
+
+#import for kstest based estimation
+#should be replace
+# FIXME: importing these patches scipy distribution classes in-place.
+#  Don't do this.
+import statsmodels.sandbox.distributions.sppatch  # noqa:F401
+
 
 #redefine some shortcuts
 np_log = np.log
@@ -137,7 +145,6 @@ def nloglikeobs(self, params):
 print(resp2.params)
 print(resp2.bse)
 
-from statsmodels.tools.numdiff import approx_hess
 
 hb=-approx_hess(modp.start_params, modp.loglike, epsilon=-1e-4)
 tmp = modp.loglike(modp.start_params)
@@ -156,15 +163,6 @@ def nloglikeobs(self, params):
 # fit_ks works well, but no bse or other result statistics yet
 
 
-
-
-
-#import for kstest based estimation
-#should be replace
-# FIXME: importing these patches scipy distribution classes in-place.
-#  Don't do this.
-import statsmodels.sandbox.distributions.sppatch  # noqa:F401
-
 class MyPareto(GenericLikelihoodModel):
     '''Maximum Likelihood Estimation pareto distribution
 

statsmodels/examples/ex_kernel_regression.py

@@ -9,6 +9,7 @@
 from __future__ import print_function
 import numpy as np
 import numpy.testing as npt
+import matplotlib.pyplot as plt
 
 import statsmodels.nonparametric.api as nparam
 #import statsmodels.api as sm
@@ -54,7 +55,6 @@
 npt.assert_allclose(sm_mean2, R_mean, atol=1e-2)
 npt.assert_allclose(sm_R2, R_R2, atol=1e-2)
 
-import matplotlib.pyplot as plt
 
 fig = plt.figure()
 ax = fig.add_subplot(1,1,1)

statsmodels/examples/ex_lowess.py

@@ -8,16 +8,19 @@
 """
 
 from __future__ import print_function
+import os
+
 import numpy as np
 import matplotlib.pyplot as plt
 import statsmodels.api as sm
-
-lowess = sm.nonparametric.lowess
+import statsmodels.nonparametric.tests.results
 
 # this is just to check direct import
 import statsmodels.nonparametric.smoothers_lowess
 statsmodels.nonparametric.smoothers_lowess.lowess
 
+lowess = sm.nonparametric.lowess
+
 x = np.arange(20.)
 
 #standard normal noise
@@ -60,8 +63,6 @@
 plt.plot(actual_lowess[:,1])
 plt.plot(expected_lowess[:,1])
 
-import os.path
-import statsmodels.nonparametric.tests.results
 rpath = os.path.split(statsmodels.nonparametric.tests.results.__file__)[0]
 rfile = os.path.join(rpath, 'test_lowess_frac.csv')
 test_data = np.genfromtxt(open(rfile, 'rb'),

statsmodels/examples/ex_misc_tarma.py

@@ -8,10 +8,12 @@
 
 from __future__ import print_function
 import numpy as np
+import matplotlib.pyplot as plt
 
 from statsmodels.tsa.arima_process import arma_generate_sample, ArmaProcess
 from statsmodels.miscmodels.tmodel import TArma
 from statsmodels.tsa.arima_model import ARMA
+from statsmodels.tsa.arma_mle import Arma
 
 nobs = 500
 ar = [1, -0.6, -0.1]
@@ -36,7 +38,6 @@
 
 print(proc.ar_roots(), proc.ma_roots())
 
-from statsmodels.tsa.arma_mle import Arma
 modn = Arma(x)
 resn = modn.fit_mle(order=order)
 
@@ -69,7 +70,6 @@
 resid = res2.model.geterrors(res2.params)
 fv = res[2]['fvec']  #resid returned from leastsq?
 
-import matplotlib.pyplot as plt
 plt.plot(x, 'o', alpha=0.5)
 plt.plot(x-resid)
 plt.plot(x-fv)

statsmodels/examples/ex_misc_tmodel.py

@@ -5,6 +5,7 @@
 from scipy import stats
 import statsmodels.api as sm
 from statsmodels.miscmodels import TLinearModel
+from statsmodels.tools.numdiff import approx_hess
 
 #Example:
 #np.random.seed(98765678)
@@ -69,7 +70,6 @@
 print(resp2.params)
 print(resp2.bse)
 
-from statsmodels.tools.numdiff import approx_hess
 
 hb=-approx_hess(modp.start_params, modp.loglike, epsilon=-1e-4)
 tmp = modp.loglike(modp.start_params)

statsmodels/examples/ex_pandas.py

@@ -10,11 +10,14 @@
 from datetime import datetime
 
 import numpy as np
+import matplotlib.pyplot as plt
+import matplotlib as mpl
 
 from pandas import DataFrame, Series
 
 import statsmodels.api as sm
 import statsmodels.tsa.api as tsa
+from statsmodels.tsa.arima_process import arma_generate_sample
 
 
 data = sm.datasets.stackloss.load(as_pandas=False)
@@ -43,9 +46,6 @@
 print(hub_results.summary())
 
 
-import matplotlib.pyplot as plt
-import matplotlib as mpl
-
 def plot_acf_multiple(ys, lags=20):
     """
 
@@ -109,12 +109,8 @@ def plot_acf_multiple(ys, lags=20):
 
 #Example TSA ARMA
 
-import numpy as np
-import statsmodels.api as sm
-
 
 # Generate some data from an ARMA process
-from statsmodels.tsa.arima_process import arma_generate_sample
 arparams = np.array([.75, -.25])
 maparams = np.array([.65, .35])
 # The conventions of the arma_generate function require that we specify a

statsmodels/examples/ex_predict_results.py

@@ -8,19 +8,22 @@
 """
 
 import numpy as np
+from numpy.testing import assert_allclose
 from statsmodels.regression.linear_model import WLS
 
+from statsmodels.tools.tools import add_constant
 from statsmodels.sandbox.regression.predstd import wls_prediction_std
 from statsmodels.regression._prediction import get_prediction
 from statsmodels.genmod._prediction import params_transform_univariate
+from statsmodels.genmod.generalized_linear_model import GLM
+from statsmodels.genmod.families import links
 
 
 # from example wls.py
 
 nsample = 50
 x = np.linspace(0, 20, nsample)
 X = np.column_stack((x, (x - 5)**2))
-from statsmodels.tools.tools import add_constant
 X = add_constant(X)
 beta = [5., 0.5, -0.01]
 sig = 0.5
@@ -43,7 +46,6 @@
 pred_res = get_prediction(res_wls)
 ci = pred_res.conf_int(obs=True)
 
-from numpy.testing import assert_allclose
 assert_allclose(pred_res.se_obs, prstd, rtol=1e-13)
 assert_allclose(ci, np.column_stack((iv_l, iv_u)), rtol=1e-13)
 
@@ -52,7 +54,6 @@
 pred_res2 = res_wls.get_prediction()
 ci2 = pred_res2.conf_int(obs=True)
 
-from numpy.testing import assert_allclose
 assert_allclose(pred_res2.se_obs, prstd, rtol=1e-13)
 assert_allclose(ci2, np.column_stack((iv_l, iv_u)), rtol=1e-13)
 
@@ -62,7 +63,6 @@
 pred_wls_n = res_wls_n.get_prediction()
 print(pred_wls_n.summary_frame().head())
 
-from statsmodels.genmod.generalized_linear_model import GLM
 
 w_sqrt = np.sqrt(w)
 mod_glm = GLM(y/w_sqrt, X/w_sqrt[:,None])
@@ -83,7 +83,6 @@
                           np.exp(res_glm.conf_int())))
 assert_allclose(rates.summary_frame().values, rates2, rtol=1e-13)
 
-from statsmodels.genmod.families import links
 
 # with identity transform
 pt = params_transform_univariate(res_glm.params, res_glm.cov_params(), link=links.identity())

statsmodels/examples/ex_proportion.py

@@ -9,6 +9,8 @@
 from __future__ import print_function
 from statsmodels.compat.python import lmap
 import numpy as np
+import matplotlib.pyplot as plt
+
 import statsmodels.stats.proportion as sms
 import statsmodels.stats.weightstats as smw
 
@@ -45,7 +47,6 @@
 nn = np.arange(200, 351)
 pow_z = sms.power_ztost_prop(0.5, 0.72, nn, 0.6, alpha=0.05)
 pow_bin = sms.power_ztost_prop(0.5, 0.72, nn, 0.6, alpha=0.05, dist='binom')
-import matplotlib.pyplot as plt
 plt.plot(nn, pow_z[0], label='normal')
 plt.plot(nn, pow_bin[0], label='binomial')
 plt.legend(loc='lower right')

statsmodels/examples/ex_regressionplots.py

@@ -12,6 +12,7 @@
 
 from statsmodels.sandbox.regression.predstd import wls_prediction_std
 import statsmodels.graphics.regressionplots as smrp
+from statsmodels.graphics.tests.test_regressionplots import TestPlot
 
 #example from tut.ols with changes
 #fix a seed for these examples
@@ -122,7 +123,6 @@
     smrp.plot_ccpr(res, exog_idx=0)
     smrp.plot_ccpr_grid(res, exog_idx=[0,1])
 
-from statsmodels.graphics.tests.test_regressionplots import TestPlot
 tp = TestPlot()
 tp.test_plot_fit()
 

statsmodels/examples/ex_wald_anova.py

@@ -14,6 +14,8 @@
 from statsmodels.discrete.discrete_model import Poisson
 
 import statsmodels.stats.tests.test_anova as ttmod
+from statsmodels.discrete.discrete_model import NegativeBinomial
+
 
 test = ttmod.TestAnova3()
 test.setup_class()
@@ -36,7 +38,6 @@
 print('\nPoisson 2')
 print(res_poi_2.wald_test_terms(skip_single=False))
 
-from statsmodels.discrete.discrete_model import NegativeBinomial
 res_nb2 = NegativeBinomial.from_formula("Days ~ C(Weight) * C(Duration)", data).fit()
 print('\nNegative Binomial nb2')
 print(res_nb2.wald_test_terms(skip_single=False))

statsmodels/examples/example_discrete_mnl.py

@@ -5,6 +5,9 @@
 from statsmodels.compat.python import lrange
 import numpy as np
 import statsmodels.api as sm
+from statsmodels.iolib.summary import (
+    table_extend, summary_params_2d, summary_params_2dflat)
+
 
 anes_data = sm.datasets.anes96.load(as_pandas=False)
 anes_exog = anes_data.exog
@@ -20,9 +23,6 @@
 #mlogit_res = mlogit_mod.fit(method='ncg') # this takes forever
 
 
-from statsmodels.iolib.summary import (
-                        summary_params_2d, summary_params_2dflat)
-
 exog_names = [anes_data.exog_name[i] for i in [0, 2]+lrange(5,8)] + ['const']
 endog_names = [anes_data.endog_name+'_%d' % i for i in np.unique(mlogit_res.model.endog)[1:]]
 print('\n\nMultinomial')
@@ -41,7 +41,6 @@
 print('\n\n')
 print('\n'.join((str(t) for t in tables)))
 
-from statsmodels.iolib.summary import table_extend
 at = table_extend(tables)
 print(at)