Mac OS 10.8 Tests

130 views
Skip to first unread message

Tom Augspurger

unread,
Jun 26, 2013, 9:44:09 AM6/26/13
to pystat...@googlegroups.com
In preparation for the 0.5 release:

In [6]: statsmodels.__version__
Out[6]: '0.5.0.dev-917da60'


(python2.7)Toms-MacBook-Pro:python2.7 tom$ python -c "import statsmodels as s; 
s.test('full')"
Running unit tests for statsmodels
NumPy version 1.8.0.dev-b5fd429
NumPy is installed in /Users/tom/python2.7/lib/python2.7/site-packages/numpy
Python version 2.7.3 (default, Feb 10 2013, 17:17:55) [GCC 4.2.1 Compatible Apple LLVM 4.2 (clang-425.0.24)]
nose version 1.3.0
/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/vector_ar/var_model.py:339: FutureWarning: The names argument is deprecated and will be removed in the next release.
  "removed in the next release.", FutureWarning)
.........................................................................................................................S.....S.......................................................................S..............................................................................................................E...........................................................FFFFFFF......................................................................................QC check did not pass for 3 out of 4 parameters
Try increasing solver accuracy or number of iterations, decreasing alpha, or switch solvers
F..................EEEEE....................................................................................................................................................................................................................................EE..................................E.E............................................................S..............................................................................................................................................................................................................................................................................................S........S..........S........S.........S..........S........S.........S.........S...........................................................................................E....................../Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/stats/correlation_tools.py:82: UserWarning: maximum iteration reached
  warnings.warn('maximum iteration reached')
.......................................................................................................................................................................................................................................................................................................................................................................................................................................................S....................................................................F.........no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
..no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
......./Users/tom/python2.7/lib/python2.7/site-packages/scipy/stats/stats.py:1292: UserWarning: kurtosistest only valid for n>=20 ... continuing anyway, n=15
  int(n))
..................................................................................................................................................................................................F..................EEE.F............................................................................EEEEEEEE............................................................EEEE................................................................EEEE.E.E...EEEEE..EE.EE.........................................................
======================================================================
ERROR: test suite for <class 'statsmodels.discrete.tests.test_discrete.TestPoissonL1Compatability'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 562, in setupClass
    trim_mode='auto')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/discrete_model.py", line 756, in fit_regularized
    size_trim_tol=size_trim_tol, qc_tol=qc_tol, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/discrete_model.py", line 312, in fit_regularized
    cov_params_func=cov_params_func, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/base/model.py", line 330, in fit
    Hinv = cov_params_func(self, xopt, retvals)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/discrete_model.py", line 332, in cov_params_func_l1
    H_restricted_inv = np.linalg.inv(-H_restricted)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/linalg/linalg.py", line 500, in inv
    ainv = _umath_linalg.inv(a, signature=signature, extobj=extobj)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/linalg/linalg.py", line 88, in _raise_linalgerror_singular
    raise LinAlgError("Singular matrix")
LinAlgError: Singular matrix

======================================================================
ERROR: Failure: IndexError (too many indices)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 518, in makeTest
    return self._makeTest(obj, parent)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 577, in _makeTest
    return MethodTestCase(obj)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 345, in __init__
    self.inst = self.cls()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 19, in __init__
    super(Test_AFTModel, self).__init__()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 12, in __init__
    self.mod1 = sm.emplike.emplikeAFT(endog, exog, data.censors)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/aft_el.py", line 248, in __init__
    self.uncens_endog = self.endog[np.bool_(self.censors), :].\
IndexError: too many indices

======================================================================
ERROR: Failure: IndexError (too many indices)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 518, in makeTest
    return self._makeTest(obj, parent)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 577, in _makeTest
    return MethodTestCase(obj)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 345, in __init__
    self.inst = self.cls()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 19, in __init__
    super(Test_AFTModel, self).__init__()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 12, in __init__
    self.mod1 = sm.emplike.emplikeAFT(endog, exog, data.censors)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/aft_el.py", line 248, in __init__
    self.uncens_endog = self.endog[np.bool_(self.censors), :].\
IndexError: too many indices

======================================================================
ERROR: Failure: IndexError (too many indices)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 518, in makeTest
    return self._makeTest(obj, parent)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 577, in _makeTest
    return MethodTestCase(obj)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 345, in __init__
    self.inst = self.cls()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 19, in __init__
    super(Test_AFTModel, self).__init__()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 12, in __init__
    self.mod1 = sm.emplike.emplikeAFT(endog, exog, data.censors)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/aft_el.py", line 248, in __init__
    self.uncens_endog = self.endog[np.bool_(self.censors), :].\
IndexError: too many indices

======================================================================
ERROR: Failure: IndexError (too many indices)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 518, in makeTest
    return self._makeTest(obj, parent)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 577, in _makeTest
    return MethodTestCase(obj)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 345, in __init__
    self.inst = self.cls()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 19, in __init__
    super(Test_AFTModel, self).__init__()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 12, in __init__
    self.mod1 = sm.emplike.emplikeAFT(endog, exog, data.censors)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/aft_el.py", line 248, in __init__
    self.uncens_endog = self.endog[np.bool_(self.censors), :].\
IndexError: too many indices

======================================================================
ERROR: Failure: IndexError (too many indices)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 518, in makeTest
    return self._makeTest(obj, parent)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/loader.py", line 577, in _makeTest
    return MethodTestCase(obj)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 345, in __init__
    self.inst = self.cls()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 19, in __init__
    super(Test_AFTModel, self).__init__()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/tests/test_aft.py", line 12, in __init__
    self.mod1 = sm.emplike.emplikeAFT(endog, exog, data.censors)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/aft_el.py", line 248, in __init__
    self.uncens_endog = self.endog[np.bool_(self.censors), :].\
IndexError: too many indices

======================================================================
ERROR: statsmodels.graphics.tests.test_correlation.test_plot_corr
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/decorators.py", line 149, in skipper_func
    return f(*args, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/tests/test_correlation.py", line 20, in test_plot_corr
    fig = plot_corr(corr_matrix, xnames=hie_data.names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/correlation.py", line 115, in plot_corr
    fig.tight_layout()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/figure.py", line 1543, in tight_layout
    rect=rect)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 352, in get_tight_layout_figure
    pad=pad, h_pad=h_pad, w_pad=w_pad)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 129, in auto_adjust_subplotpars
    tight_bbox_raw = union([ax.get_tightbbox(renderer) for ax in subplots])
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axes.py", line 8895, in get_tightbbox
    bb.append(self.title.get_window_extent(renderer))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 752, in get_window_extent
    bbox, info = self._get_layout(self._renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 304, in _get_layout
    ismath=False)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/backends/backend_macosx.py", line 159, in get_text_width_height_descent
    width, height, descent = self.gc.get_text_width_height_descent(unicode(s), family, size, weight, style)
RuntimeError: CGContextRef is NULL

======================================================================
ERROR: statsmodels.graphics.tests.test_correlation.test_plot_corr_grid
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/decorators.py", line 149, in skipper_func
    return f(*args, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/tests/test_correlation.py", line 35, in test_plot_corr_grid
    fig = plot_corr_grid([corr_matrix] * 2, xnames=hie_data.names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/correlation.py", line 218, in plot_corr_grid
    normcolor=normcolor, ax=ax, cmap=cmap)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/correlation.py", line 115, in plot_corr
    fig.tight_layout()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/figure.py", line 1543, in tight_layout
    rect=rect)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 352, in get_tight_layout_figure
    pad=pad, h_pad=h_pad, w_pad=w_pad)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 129, in auto_adjust_subplotpars
    tight_bbox_raw = union([ax.get_tightbbox(renderer) for ax in subplots])
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axes.py", line 8897, in get_tightbbox
    bb_xaxis = self.xaxis.get_tightbbox(renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axis.py", line 1018, in get_tightbbox
    renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axis.py", line 1001, in _get_tick_bboxes
    extent = tick.label1.get_window_extent(renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 752, in get_window_extent
    bbox, info = self._get_layout(self._renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 304, in _get_layout
    ismath=False)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/backends/backend_macosx.py", line 159, in get_text_width_height_descent
    width, height, descent = self.gc.get_text_width_height_descent(unicode(s), family, size, weight, style)
RuntimeError: CGContextRef is NULL

======================================================================
ERROR: statsmodels.graphics.tests.test_regressionplots.TestPlot.test_plot_oth
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/tests/test_regressionplots.py", line 72, in test_plot_oth
    plot_partregress_grid(res, exog_idx=[0,1])
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/regressionplots.py", line 466, in plot_partregress_grid
    fig = utils.maybe_tight_layout(fig)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/utils.py", line 156, in maybe_tight_layout
    fig.tight_layout()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/figure.py", line 1543, in tight_layout
    rect=rect)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 352, in get_tight_layout_figure
    pad=pad, h_pad=h_pad, w_pad=w_pad)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 129, in auto_adjust_subplotpars
    tight_bbox_raw = union([ax.get_tightbbox(renderer) for ax in subplots])
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axes.py", line 8897, in get_tightbbox
    bb_xaxis = self.xaxis.get_tightbbox(renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axis.py", line 1018, in get_tightbbox
    renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axis.py", line 1001, in _get_tick_bboxes
    extent = tick.label1.get_window_extent(renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 752, in get_window_extent
    bbox, info = self._get_layout(self._renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 304, in _get_layout
    ismath=False)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/backends/backend_macosx.py", line 159, in get_text_width_height_descent
    width, height, descent = self.gc.get_text_width_height_descent(unicode(s), family, size, weight, style)
RuntimeError: CGContextRef is NULL

======================================================================
ERROR: statsmodels.graphics.tests.test_regressionplots.TestPlotPandas.test_plot_oth
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/tests/test_regressionplots.py", line 72, in test_plot_oth
    plot_partregress_grid(res, exog_idx=[0,1])
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/regressionplots.py", line 466, in plot_partregress_grid
    fig = utils.maybe_tight_layout(fig)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/graphics/utils.py", line 156, in maybe_tight_layout
    fig.tight_layout()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/figure.py", line 1543, in tight_layout
    rect=rect)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 352, in get_tight_layout_figure
    pad=pad, h_pad=h_pad, w_pad=w_pad)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/tight_layout.py", line 129, in auto_adjust_subplotpars
    tight_bbox_raw = union([ax.get_tightbbox(renderer) for ax in subplots])
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axes.py", line 8897, in get_tightbbox
    bb_xaxis = self.xaxis.get_tightbbox(renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axis.py", line 1018, in get_tightbbox
    renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/axis.py", line 1001, in _get_tick_bboxes
    extent = tick.label1.get_window_extent(renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 752, in get_window_extent
    bbox, info = self._get_layout(self._renderer)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/text.py", line 304, in _get_layout
    ismath=False)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/backends/backend_macosx.py", line 159, in get_text_width_height_descent
    width, height, descent = self.gc.get_text_width_height_descent(unicode(s), family, size, weight, style)
RuntimeError: CGContextRef is NULL

======================================================================
ERROR: test_10000_to_ch (statsmodels.stats.libqsturng.tests.test_qsturng.test_qsturng)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/stats/libqsturng/tests/test_qsturng.py", line 123, in test_10000_to_ch
    ps, rs, vs, qs = read_ch(curdir + '/bootleg.dat') # <- generated by qtukey in R
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/stats/libqsturng/tests/test_qsturng.py", line 19, in read_ch
    with open(fname) as f:
IOError: [Errno 2] No such file or directory: '/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/stats/libqsturng/tests/bootleg.dat'

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_ARIMA101'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 544, in setupClass
    cls.res1 = ARIMA(endog, (1,0,1)).fit(trend="c", disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_ARIMA111'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 557, in setupClass
    cls.res1 = ARIMA(cpi, (1,1,1)).fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_ARIMA111CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 587, in setupClass
    cls.res1 = ARIMA(cpi, (1,1,1)).fit(disp=-1, method='css')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA11_Const'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 263, in setupClass
    cls.res1 = ARMA(endog, order=(1,1)).fit(trend="c", disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA11_Const_CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 406, in setupClass
    disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA11_NoConst'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 197, in setupClass
    cls.res1 = ARMA(endog, order=(1,1)).fit(trend='nc', disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA11_NoConst_CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 327, in setupClass
    disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA14_Const'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 273, in setupClass
    cls.res1 = ARMA(endog, order=(1,4)).fit(trend="c", disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA14_Const_CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 418, in setupClass
    disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA14_NoConst'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 216, in setupClass
    cls.res1 = ARMA(endog, order=(1,4)).fit(trend='nc', disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA14_NoConst_CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 337, in setupClass
    disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA41_Const'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 282, in setupClass
    cls.res1 = ARMA(endog, order=(4,1)).fit(trend="c", disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA41_Const_CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 429, in setupClass
    disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA41_NoConst'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 226, in setupClass
    cls.res1 = ARMA(endog, order=(4,1)).fit(trend='nc', disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_Y_ARMA41_NoConst_CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 352, in setupClass
    disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_compare_arma
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 65, in test_compare_arma
    reskf = modkf.fit(trend='nc', disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_reset_trend
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 471, in test_reset_trend
    res1 = mod.fit(trend="c", disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_start_params_bug
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 537, in test_start_params_bug
    res = ARMA(data, order=(4,1)).fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_mle_dates
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 673, in test_arima_predict_mle_dates
    res1 = ARIMA(cpi, (4,1,1), dates=cpi_dates, freq='Q').fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_css_dates
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 726, in test_arima_predict_css_dates
    method='css', trend='nc')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_mle
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 773, in test_arima_predict_mle
    res1 = ARIMA(cpi, (4,1,1)).fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_css
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1147, in test_arima_predict_css
    trend="nc")
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_css_diffs
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1286, in test_arima_predict_css_diffs
    trend="c")
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_mle_diffs
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1429, in test_arima_predict_mle_diffs
    res1 = ARIMA(cpi, (4,1,1)).fit(disp=-1, trend="c")
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_wrapper
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1567, in test_arima_wrapper
    res = ARIMA(cpi, (4,1,1), freq='Q').fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_1dexog
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1579, in test_1dexog
    mod = ARMA(endog, (1,1), exog).fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_pandas_nofreq
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1616, in test_arima_predict_pandas_nofreq
    arma = ARMA(data, order=(1,0)).fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_exog
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1669, in test_arima_predict_exog
    disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_noma
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1727, in test_arima_predict_noma
    arma_res = arma.fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arimax
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1749, in test_arimax
    ftol=1e-12, xtol=1e-12)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 997, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_aic
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 394, in test_aic
    self.res1.aic, self.res2.aic, DECIMAL_3)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 474, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 3 decimals
 ACTUAL: 44.36141955583647
 DESIRED: 38.39977387754293

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_bic
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 398, in test_bic
    self.res1.bic, self.res2.bic, DECIMAL_3)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 474, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 3 decimals
 ACTUAL: 44.36141955583647
 DESIRED: 42.796981585942106

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_bse
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 386, in test_bse
    assert_almost_equal(self.res1.bse, self.res2.bse, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 615, in assert_array_compare
    chk_same_position(x_isnan, y_isnan, hasval='nan')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 595, in chk_same_position
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

x and y nan location mismatch:
 x: array([ nan,  nan,  nan,  nan])
 y: array([ 2.05922641,  0.61889778,  0.07383875,         nan])

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_conf_int
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 383, in test_conf_int
    self.res1.conf_int(), self.res2.conf_int, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 615, in assert_array_compare
    chk_same_position(x_isnan, y_isnan, hasval='nan')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 595, in chk_same_position
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

x and y nan location mismatch:
 x: array([[ nan,  nan],
       [ nan,  nan],
       [ nan,  nan],
       [ nan,  nan]])
 y: array([[-9.44077951, -1.36876033],
       [ 0.03716721,  2.46320194],
       [-0.09727571,  0.19216687],
       [        nan,         nan]])

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_cov_params
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 416, in test_cov_params
    self.res1.cov_params(), self.res2.cov_params, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 615, in assert_array_compare
    chk_same_position(x_isnan, y_isnan, hasval='nan')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 595, in chk_same_position
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

x and y nan location mismatch:
 x: array([[ nan,  nan,  nan,  nan],
       [ nan,  nan,  nan,  nan],
       [ nan,  nan,  nan,  nan],
       [ nan,  nan,  nan,  nan]])
 y: array([[ 4.24041339, -0.83432592, -0.06827915,         nan],
       [-0.83432592,  0.38303447, -0.01700249,         nan],
       [-0.06827915, -0.01700249,  0.00545216,         nan],
       [        nan,         nan,         nan,         nan]])

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_nnz_params
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 390, in test_nnz_params
    self.res1.nnz_params, self.res2.nnz_params, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 474, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals
 ACTUAL: 0
 DESIRED: 3

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_params
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 379, in test_params
    assert_almost_equal(self.res1.params, self.res2.params, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 652, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

(mismatch 75.0%)
 x: array([ 0.,  0.,  0.,  0.])
 y: array([-5.40476992,  1.25018458,  0.04744558,  0.        ])

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestSweepAlphaL1.test_sweep_alpha
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 494, in test_sweep_alpha
    assert_almost_equal(res2.params, self.res1.params[i], DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 652, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

(mismatch 100.0%)
 x: array([-4163.47806032,  1019.39255645,     6.04894582,   902.32935339])
 y: array([-10.37593611,   2.27080968,   0.06670638,   2.05723691])

======================================================================
FAIL: statsmodels.stats.tests.test_power.test_power_solver_warn
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/stats/tests/test_power.py", line 725, in test_power_solver_warn
    assert_almost_equal(val, 1600, decimal=4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 474, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals
 ACTUAL: 6.9030809341711435e-06
 DESIRED: 1600

======================================================================
FAIL: statsmodels.tsa.tests.test_ar.TestARMLEConstant.test_dynamic_predict
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_ar.py", line 159, in test_dynamic_predict
    assert_almost_equal(fv, res2.fcdyn[start:end+1], DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 652, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

(mismatch 27.9069767442%)
 x: array([ 11.23049853,  13.23737915,  19.66416987,  27.00383112,
        33.57868265,  44.90793235,  47.95875227,  45.09355264,
        36.50664185,  28.78826358,  23.48721623,  22.53218482,...
 y: array([ 11.23047,  13.23731,  19.66407,  27.00373,  33.57858,  44.90784,
        47.95866,  45.09347,  36.50656,  28.78818,  23.48711,  22.53206,
        25.68737,  31.07822,  38.54836,  45.41845,  49.67139,  49.37283,...

======================================================================
FAIL: statsmodels.tsa.tests.test_arima.Test_ARIMA112CSS.test_arroots
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 105, in test_arroots
    self.decimal_arroots)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 454, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 3 decimals
 ACTUAL: array([ -3.72917037e-155+0.j])
 DESIRED: [(-1.4442+0j)]

----------------------------------------------------------------------
Ran 2315 tests in 456.576s

FAILED (SKIP=14, errors=41, failures=11)

I'd be happy to run more tests today.  I've also got a python 3 environment setup.  I can run those as well if you'd like.

josef...@gmail.com

unread,
Jun 26, 2013, 9:55:45 AM6/26/13
to pystat...@googlegroups.com
On Wed, Jun 26, 2013 at 9:44 AM, Tom Augspurger
<tom.augs...@gmail.com> wrote:
> In preparation for the 0.5 release:
>
> In [6]: statsmodels.__version__
> Out[6]: '0.5.0.dev-917da60'
>
>
> (python2.7)Toms-MacBook-Pro:python2.7 tom$ python -c "import statsmodels as
> s;
> s.test('full')"
> Running unit tests for statsmodels
> NumPy version 1.8.0.dev-b5fd429

preliminary question given some of the errors

Are you using development master versions for both numpy and scipy ?
Can you also test against the released version on the same machine, so
we have a diff to what's our problems and what has changed in numpy
and scipy?

For example the L1 failures look like a recent change in scipy master
(possibly a bug in scipy, or we need to adjust our code)

I have not tested anything against scipy and numpy master since I
cannot compile them.

Josef

josef...@gmail.com

unread,
Jun 26, 2013, 9:59:18 AM6/26/13
to pystat...@googlegroups.com
On Wed, Jun 26, 2013 at 9:55 AM, <josef...@gmail.com> wrote:
> On Wed, Jun 26, 2013 at 9:44 AM, Tom Augspurger
> <tom.augs...@gmail.com> wrote:
>> In preparation for the 0.5 release:
>>
>> In [6]: statsmodels.__version__
>> Out[6]: '0.5.0.dev-917da60'

This might be an outdated version. I need to verify the hash.

Josef

josef...@gmail.com

unread,
Jun 26, 2013, 10:03:25 AM6/26/13
to pystat...@googlegroups.com
On Wed, Jun 26, 2013 at 9:59 AM, <josef...@gmail.com> wrote:
> On Wed, Jun 26, 2013 at 9:55 AM, <josef...@gmail.com> wrote:
>> On Wed, Jun 26, 2013 at 9:44 AM, Tom Augspurger
>> <tom.augs...@gmail.com> wrote:
>>> In preparation for the 0.5 release:
>>>
>>> In [6]: statsmodels.__version__
>>> Out[6]: '0.5.0.dev-917da60'
>
> This might be an outdated version. I need to verify the hash.

https://github.com/statsmodels/statsmodels/commits/master

It looks like you took an (not up-to-date) master from my fork
josef-pkt/statsmodels instead of statsmodels/statsmodels.

Josef

Tom Augspurger

unread,
Jun 26, 2013, 10:07:11 AM6/26/13
to pystat...@googlegroups.com


On Wednesday, June 26, 2013 9:03:25 AM UTC-5, josefpktd wrote:
On Wed, Jun 26, 2013 at 9:59 AM,  <josef...@gmail.com> wrote:
> On Wed, Jun 26, 2013 at 9:55 AM,  <josef...@gmail.com> wrote:
>> On Wed, Jun 26, 2013 at 9:44 AM, Tom Augspurger
>> <tom.augs...@gmail.com> wrote:
>>> In preparation for the 0.5 release:
>>>
>>> In [6]: statsmodels.__version__
>>> Out[6]: '0.5.0.dev-917da60'
>
> This might be an outdated version. I need to verify the hash.

https://github.com/statsmodels/statsmodels/commits/master

It looks like you took an (not up-to-date) master from my fork
josef-pkt/statsmodels instead of statsmodels/statsmodels.

Odd... I'll get that fixed and rerun.  If those still fail/error I'll get setup a virtualenv and test against the stable versions of numpy/scipy. 

josef...@gmail.com

unread,
Jun 26, 2013, 10:19:47 AM6/26/13
to pystat...@googlegroups.com
On Wed, Jun 26, 2013 at 10:07 AM, Tom Augspurger
<tom.augs...@gmail.com> wrote:
>
>
> On Wednesday, June 26, 2013 9:03:25 AM UTC-5, josefpktd wrote:
>>
>> On Wed, Jun 26, 2013 at 9:59 AM, <josef...@gmail.com> wrote:
>> > On Wed, Jun 26, 2013 at 9:55 AM, <josef...@gmail.com> wrote:
>> >> On Wed, Jun 26, 2013 at 9:44 AM, Tom Augspurger
>> >> <tom.augs...@gmail.com> wrote:
>> >>> In preparation for the 0.5 release:
>> >>>
>> >>> In [6]: statsmodels.__version__
>> >>> Out[6]: '0.5.0.dev-917da60'
>> >
>> > This might be an outdated version. I need to verify the hash.
>>
>> https://github.com/statsmodels/statsmodels/commits/master
>>
>> It looks like you took an (not up-to-date) master from my fork
>> josef-pkt/statsmodels instead of statsmodels/statsmodels.
>>
> Odd... I'll get that fixed and rerun. If those still fail/error I'll get
> setup a virtualenv and test against the stable versions of numpy/scipy.

Thank you, that would be very helpful.

Looking at your test failures: which matplotlib version are you using?

two recurrent errors

boolean indexing in numpy might have changed

======================================================================
ERROR: Failure: IndexError (too many indices)
----------------------------------------------------------------------
Traceback (most recent call last):
<...>
File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/emplike/aft_el.py",
line 248, in __init__
self.uncens_endog = self.endog[np.bool_(self.censors), :].\
IndexError: too many indices


matplotlib backend errors that I have never seen before

======================================================================
ERROR: statsmodels.graphics.tests.test_correlation.test_plot_corr
----------------------------------------------------------------------
Traceback (most recent call last):
<...>
File "/Users/tom/python2.7/lib/python2.7/site-packages/matplotlib/backends/backend_macosx.py",
line 159, in get_text_width_height_descent
width, height, descent =
self.gc.get_text_width_height_descent(unicode(s), family, size,
weight, style)
RuntimeError: CGContextRef is NULL


some of the other errors and failures are already fixed in statsmodels.master

The ARMA failures might be difficult to debug, since we don't get the
exception where the problem actually is.

Josef

Tom Augspurger

unread,
Jun 26, 2013, 10:24:27 AM6/26/13
to pystat...@googlegroups.com


On Wednesday, June 26, 2013 9:07:11 AM UTC-5, Tom Augspurger wrote:
 
Odd... I'll get that fixed and rerun.  If those still fail/error I'll get setup a virtualenv and test against the stable versions of numpy/scipy. 

New results:

In [1]: import statsmodels; statsmodels.__version__
Out[1]: '0.5.0.dev-947943d'
 
(python2.7)Toms-MacBook-Pro:python2.7 tom$ python -c "import statsmodels as s; s.test('full')"
Running unit tests for statsmodels
NumPy version 1.8.0.dev-b5fd429
NumPy is installed in /Users/tom/python2.7/lib/python2.7/site-packages/numpy
Python version 2.7.3 (default, Feb 10 2013, 17:17:55) [GCC 4.2.1 Compatible Apple LLVM 4.2 (clang-425.0.24)]
nose version 1.3.0
/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/vector_ar/var_model.py:339: FutureWarning: The names argument is deprecated and will be removed in the next release.
  "removed in the next release.", FutureWarning)
.........................................................................................................................S.....S.......................................................................S..............................................................................................................E...........................................................FFFFFFF......................................................................................QC check did not pass for 3 out of 4 parameters
Try increasing solver accuracy or number of iterations, decreasing alpha, or switch solvers
F..................EEEEE..................................................................................................................................................................................................................................................EE..................................E.E............................................................S.................................................................................................................................................................................................................................................................................................................................................................................................................................................................................S..........S........S...S..........S........S..S........S...S..........S........S..S.......S...S..................................................................................................................../Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/stats/correlation_tools.py:82: UserWarning: maximum iteration reached
  warnings.warn('maximum iteration reached')
.......................................................................................................................................................................................................................................................................................................................................................................................................................................................S..............................................................................no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
..no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
no overlap, power is zero, TODO
......./Users/tom/python2.7/lib/python2.7/site-packages/scipy/stats/stats.py:1292: UserWarning: kurtosistest only valid for n>=20 ... continuing anyway, n=15
  int(n))
..................................................................................................................................................................................................F..................EEE.F....F.......................................................................EEEEEEEE............................................................EEEE................................................................EEEE.E.E...EEEEE..EE.EE.........................................................
======================================================================
ERROR: test suite for <class 'statsmodels.discrete.tests.test_discrete.TestPoissonL1Compatability'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 561, in setupClass
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 425, in _fit_start_params_hr
    raise ValueError("The computed initial AR coefficients are not "
ValueError: The computed initial AR coefficients are not stationary
You should induce stationarity, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: test suite for <class 'statsmodels.tsa.tests.test_arima.Test_ARIMA111CSS'>
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 208, in run
    self.setUp()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 291, in setUp
    self.setupContext(ancestor)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/suite.py", line 314, in setupContext
    try_run(context, names)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/util.py", line 469, in try_run
    return func()
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 587, in setupClass
    cls.res1 = ARIMA(cpi, (1,1,1)).fit(disp=-1, method='css')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_css_dates
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 726, in test_arima_predict_css_dates
    method='css', trend='nc')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_mle
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 773, in test_arima_predict_mle
    res1 = ARIMA(cpi, (4,1,1)).fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_css
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1147, in test_arima_predict_css
    trend="nc")
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_css_diffs
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1286, in test_arima_predict_css_diffs
    trend="c")
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 442, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_predict_mle_diffs
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1429, in test_arima_predict_mle_diffs
    res1 = ARIMA(cpi, (4,1,1)).fit(disp=-1, trend="c")
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
ERROR: statsmodels.tsa.tests.test_arima.test_arima_wrapper
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 1567, in test_arima_wrapper
    res = ARIMA(cpi, (4,1,1), freq='Q').fit(disp=-1)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 998, in fit
    full_output, disp, callback, **kwargs)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 821, in fit
    start_params = self._fit_start_params((k_ar,k_ma,k), method)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 446, in _fit_start_params
    start_params = self._fit_start_params_hr(order)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/arima_model.py", line 432, in _fit_start_params_hr
    raise ValueError("The computed initial MA coefficients are not "
ValueError: The computed initial MA coefficients are not invertible
You should induce invertibility, choose a different model order, or you can
pass your own start_params.

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_aic
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 393, in test_aic
    self.res1.aic, self.res2.aic, DECIMAL_3)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 474, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 3 decimals
 ACTUAL: 44.36141955583647
 DESIRED: 38.39977387754293

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_bic
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 397, in test_bic
    self.res1.bic, self.res2.bic, DECIMAL_3)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 474, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 3 decimals
 ACTUAL: 44.36141955583647
 DESIRED: 42.796981585942106

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_bse
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 385, in test_bse
    assert_almost_equal(self.res1.bse, self.res2.bse, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 615, in assert_array_compare
    chk_same_position(x_isnan, y_isnan, hasval='nan')
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 595, in chk_same_position
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

x and y nan location mismatch:
 x: array([ nan,  nan,  nan,  nan])
 y: array([ 2.05922641,  0.61889778,  0.07383875,         nan])

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_conf_int
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 382, in test_conf_int
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 415, in test_cov_params
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 389, in test_nnz_params
    self.res1.nnz_params, self.res2.nnz_params, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 474, in assert_almost_equal
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals
 ACTUAL: 0
 DESIRED: 3

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestProbitL1.test_params
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 378, in test_params
    assert_almost_equal(self.res1.params, self.res2.params, DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 652, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

(mismatch 75.0%)
 x: array([ 0.,  0.,  0.,  0.])
 y: array([-5.40476992,  1.25018458,  0.04744558,  0.        ])

======================================================================
FAIL: statsmodels.discrete.tests.test_discrete.TestSweepAlphaL1.test_sweep_alpha
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/discrete/tests/test_discrete.py", line 493, in test_sweep_alpha
    assert_almost_equal(res2.params, self.res1.params[i], DECIMAL_4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 652, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

(mismatch 100.0%)
 x: array([-4163.47806032,  1019.39255645,     6.04894582,   902.32935339])
 y: array([-10.37593611,   2.27080968,   0.06670638,   2.05723691])

======================================================================
 ACTUAL: array([ 0.24951337+0.24999953j])
 DESIRED: [(-1.4442+0j)]

======================================================================
FAIL: statsmodels.tsa.tests.test_arima.Test_ARIMA112CSS.test_freq
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/python2.7/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_arima.py", line 649, in test_freq
    assert_almost_equal(self.res1.arfreq, [0.5000], 4)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 458, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 819, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/python2.7/lib/python2.7/site-packages/numpy/testing/utils.py", line 652, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

(mismatch 100.0%)
 x: array([ 0.1251549])
 y: array([ 0.5])

----------------------------------------------------------------------
Ran 2526 tests in 501.387s

FAILED (SKIP=19, errors=40, failures=11)


Tom Augspurger

unread,
Jun 26, 2013, 10:25:48 AM6/26/13
to pystat...@googlegroups.com


On Wednesday, June 26, 2013 9:19:47 AM UTC-5, josefpktd wrote:

Looking at your test failures: which matplotlib version are you using?

Josef

In [2]: import matplotlib; matplotlib.__version__
Out[2]: '1.2.1'

I'll maybe have a chance to dig into what's going on the the ARMA stuff later tonight. 

Tom Augspurger

unread,
Jun 26, 2013, 11:09:11 AM6/26/13
to pystat...@googlegroups.com


On Wednesday, June 26, 2013 9:19:47 AM UTC-5, josefpktd wrote:
On Wed, Jun 26, 2013 at 10:07 AM, Tom Augspurger
<tom.augs...@gmail.com> wrote:
>
>
> On Wednesday, June 26, 2013 9:03:25 AM UTC-5, josefpktd wrote:
>>
>> On Wed, Jun 26, 2013 at 9:59 AM,  <josef...@gmail.com> wrote:
>> > On Wed, Jun 26, 2013 at 9:55 AM,  <josef...@gmail.com> wrote:
>> >> On Wed, Jun 26, 2013 at 9:44 AM, Tom Augspurger
>> >> <tom.augs...@gmail.com> wrote:
>> >>> In preparation for the 0.5 release:
>> >>>
>> >>> In [6]: statsmodels.__version__
>> >>> Out[6]: '0.5.0.dev-917da60'
>> >
>> > This might be an outdated version. I need to verify the hash.
>>
>> https://github.com/statsmodels/statsmodels/commits/master
>>
>> It looks like you took an (not up-to-date) master from my fork
>> josef-pkt/statsmodels instead of statsmodels/statsmodels.
>>
> Odd... I'll get that fixed and rerun.  If those still fail/error I'll get
> setup a virtualenv and test against the stable versions of numpy/scipy.

Thank you, that would be very helpful.


Okay reran with version of numpy/scipy installed from pip.  Things look better now, but I don't know if that means more work in the future.


 There's more skips, I think since I forgot to install matplotlib.

Just one failure:

FAIL: statsmodels.tsa.tests.test_ar.TestARMLEConstant.test_dynamic_predict
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/tom/vanilla/lib/python2.7/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/Users/tom/vanilla/lib/python2.7/site-packages/statsmodels-0.5.0-py2.7-macosx-10.8-x86_64.egg/statsmodels/tsa/tests/test_ar.py", line 159, in test_dynamic_predict
    assert_almost_equal(fv, res2.fcdyn[start:end+1], DECIMAL_4)
  File "/Users/tom/vanilla/lib/python2.7/site-packages/numpy/testing/utils.py", line 452, in assert_almost_equal
    return assert_array_almost_equal(actual, desired, decimal, err_msg)
  File "/Users/tom/vanilla/lib/python2.7/site-packages/numpy/testing/utils.py", line 812, in assert_array_almost_equal
    header=('Arrays are not almost equal to %d decimals' % decimal))
  File "/Users/tom/vanilla/lib/python2.7/site-packages/numpy/testing/utils.py", line 645, in assert_array_compare
    raise AssertionError(msg)
AssertionError: 
Arrays are not almost equal to 4 decimals

(mismatch 27.9069767442%)
 x: array([ 11.23049853,  13.23737915,  19.66416987,  27.00383112,
        33.57868265,  44.90793235,  47.95875227,  45.09355264,
        36.50664185,  28.78826358,  23.48721623,  22.53218482,...
 y: array([ 11.23047,  13.23731,  19.66407,  27.00373,  33.57858,  44.90784,
        47.95866,  45.09347,  36.50656,  28.78818,  23.48711,  22.53206,
        25.68737,  31.07822,  38.54836,  45.41845,  49.67139,  49.37283,...

----------------------------------------------------------------------
Ran 2762 tests in 455.463s

FAILED (SKIP=62, failures=1)

I've had precision issues for a while; I get failures due to precision when I run the scipy test suite and I've never tracked down the source of the problem, so this failure could be idiosyncratic to me.

josef...@gmail.com

unread,
Jun 26, 2013, 11:23:18 AM6/26/13
to pystat...@googlegroups.com
On Wed, Jun 26, 2013 at 11:09 AM, Tom Augspurger
This failure shows up in one of the pythonxy Ubuntu machines, but not
in the other 3.

I have no clue on this one.
My guess is to change it to assert_allclose to get relative tolerance
instead of decimals.

>
> ----------------------------------------------------------------------
> Ran 2762 tests in 455.463s
>
> FAILED (SKIP=62, failures=1)
>
> I've had precision issues for a while; I get failures due to precision when
> I run the scipy test suite and I've never tracked down the source of the
> problem, so this failure could be idiosyncratic to me.

Much better, the question for the other failures is whether those
require our adjustments or whether those are bugs in numpy, scipy and
matplotlib.

I just installed matplotlib 1.2.1 in a virtualenv, it uses the TK
default backend (looks like it from the test noise) and I don't get
any errors.

I'll ping the numpy mailing list about the boolean slicing. ARMA needs
digging before we can tell anything.

Thanks again.

Josef

josef...@gmail.com

unread,
Jun 26, 2013, 11:45:02 AM6/26/13
to pystat...@googlegroups.com
plot failure is known
https://github.com/matplotlib/matplotlib/issues/1266
http://matplotlib.1069221.n5.nabble.com/tight-layout-causes-RuntimeError-on-OSX-tt17573.html#none

solution use Agg backend, but I have no idea if we can change anything
for this case.

Josef

josef...@gmail.com

unread,
Jun 26, 2013, 3:21:35 PM6/26/13
to pystat...@googlegroups.com
fix for indexing error with numpy master is on the way. a stray
dimension and missing squeeze/ravel in our code.

ARMA _fit_start_params_hr needs debugging with numpy master which I don't have.

Josef


On Wed, Jun 26, 2013 at 10:24 AM, Tom Augspurger
<tom.augs...@gmail.com> wrote:
>
>
> On Wednesday, June 26, 2013 9:07:11 AM UTC-5, Tom Augspurger wrote:
>
>>
>> Odd... I'll get that fixed and rerun. If those still fail/error I'll get
>> setup a virtualenv and test against the stable versions of numpy/scipy.
>
>
> New results:
>
> In [1]: import statsmodels; statsmodels.__version__
> Out[1]: '0.5.0.dev-947943d'
>
> (python2.7)Toms-MacBook-Pro:python2.7 tom$ python -c "import statsmodels as
> s; s.test('full')"
> Running unit tests for statsmodels
> NumPy version 1.8.0.dev-b5fd429
> NumPy is installed in /Users/tom/python2.7/lib/python2.7/site-packages/numpy
> Python version 2.7.3 (default, Feb 10 2013, 17:17:55) [GCC 4.2.1 Compatible
> Apple LLVM 4.2 (clang-425.0.24)]
> nose version 1.3.0

Reply all
Reply to author
Forward
0 new messages