CLN: Fixes for future SciPY and pandas

Silence new warnings produced by pandas 1.0+
Ensure starting values respect builds in HoltWinters

setup.cfg

@@ -40,6 +40,7 @@ filterwarnings =
     error:recarray support has been deprecated:FutureWarning
     error:The value returned will change to a:FutureWarning
     error:The default value of lags:FutureWarning
+    error:Support for multi-dimensional:DeprecationWarning
 markers =
     example: mark a test that runs example code
     matplotlib: mark a test that requires matplotlib

statsmodels/base/data.py

@@ -587,7 +587,7 @@ def attach_dates(self, result):
         squeezed = result.squeeze()
         k_endog = np.array(self.ynames, ndmin=1).shape[0]
         if k_endog > 1 and squeezed.shape == (k_endog,):
-            squeezed = squeezed[None, :]
+            squeezed = np.asarray(squeezed)[None, :]
         # May be zero-dim, for example in the case of forecast one step in tsa
         if squeezed.ndim < 2:
             return Series(squeezed, index=self.predict_dates)

statsmodels/base/tests/test_generic_methods.py

@@ -13,6 +13,7 @@
 from statsmodels.compat.pandas import assert_series_equal, assert_index_equal
 from statsmodels.compat.platform import (PLATFORM_OSX, PLATFORM_LINUX32,
                                          PLATFORM_WIN32)
+from statsmodels.compat.scipy import SCIPY_GT_14
 
 import numpy as np
 import pandas as pd
@@ -114,7 +115,7 @@ def test_zero_constrained(self):
         assert_allclose(tvals1, res2.tvalues, rtol=tol, atol=tol)
 
         # See gh5993
-        if PLATFORM_LINUX32:
+        if PLATFORM_LINUX32 or SCIPY_GT_14:
             pvals1 = res1.pvalues[keep_index_p]
         else:
             with pytest.warns(RuntimeWarning,
@@ -264,7 +265,7 @@ def test_zero_collinear(self):
             assert_allclose(tvals1, res2.tvalues, rtol=5e-8)
 
             # See gh5993
-            if PLATFORM_LINUX32:
+            if PLATFORM_LINUX32 or SCIPY_GT_14:
                 pvals1 = res1.pvalues[keep_index_p]
             else:
                 with pytest.warns(RuntimeWarning,

statsmodels/compat/scipy.py

@@ -6,6 +6,7 @@
 SCIPY_11 = (LooseVersion(scipy.__version__) < LooseVersion('1.2.0') and
             LooseVersion(scipy.__version__) >= LooseVersion('1.1.0'))
 
+SCIPY_GT_14 = LooseVersion(scipy.__version__) >= LooseVersion('1.5')
 
 def _next_regular(target):
     """

statsmodels/graphics/factorplots.py

@@ -130,14 +130,18 @@ def interaction_plot(x, trace, response, func=np.mean, ax=None, plottype='b',
         for i, (values, group) in enumerate(plot_data.groupby(['trace'])):
             # trace label
             label = str(group['trace'].values[0])
-            ax.plot(group['x'], group['response'], color=colors[i],
+            x = np.asarray(group['x'])
+            response = np.asarray(group['response'])
+            ax.plot(x, response, color=colors[i],
                     marker=markers[i], label=label,
                     linestyle=linestyles[i], **kwargs)
     elif plottype == 'line' or plottype == 'l':
         for i, (values, group) in enumerate(plot_data.groupby(['trace'])):
             # trace label
             label = str(group['trace'].values[0])
-            ax.plot(group['x'], group['response'], color=colors[i],
+            x = np.asarray(group['x'])
+            response = np.asarray(group['response'])
+            ax.plot(x, response, color=colors[i],
                     label=label, linestyle=linestyles[i], **kwargs)
     elif plottype == 'scatter' or plottype == 's':
         for i, (values, group) in enumerate(plot_data.groupby(['trace'])):

statsmodels/graphics/regressionplots.py

@@ -406,15 +406,17 @@ def plot_partregress(endog, exog_i, exog_others, data=None,
     # all arrays or pandas-like
 
     if RHS_isemtpy:
+        endog = np.asarray(endog)
+        exog_i = np.asarray(exog_i)
         ax.plot(endog, exog_i, 'o', **kwargs)
         fitted_line = OLS(endog, exog_i).fit()
         x_axis_endog_name = 'x' if isinstance(exog_i, np.ndarray) else exog_i.name
         y_axis_endog_name = 'y' if isinstance(endog, np.ndarray) else endog.design_info.column_names[0]
     else:
         res_yaxis = OLS(endog, RHS).fit()
         res_xaxis = OLS(exog_i, RHS).fit()
-        xaxis_resid = res_xaxis.resid
-        yaxis_resid = res_yaxis.resid
+        xaxis_resid = np.asarray(res_xaxis.resid)
+        yaxis_resid = np.asarray(res_yaxis.resid)
         x_axis_endog_name = res_xaxis.model.endog_names
         y_axis_endog_name = res_yaxis.model.endog_names
         ax.plot(xaxis_resid, yaxis_resid, 'o', **kwargs)
@@ -428,11 +430,11 @@ def plot_partregress(endog, exog_i, exog_others, data=None,
     ax.set_ylabel("e(%s | X)" % y_axis_endog_name)
     ax.set_title('Partial Regression Plot', **title_kwargs)
 
-    #NOTE: if we want to get super fancy, we could annotate if a point is
-    #clicked using this widget
-    #http://stackoverflow.com/questions/4652439/
-    #is-there-a-matplotlib-equivalent-of-matlabs-datacursormode/
-    #4674445#4674445
+    # NOTE: if we want to get super fancy, we could annotate if a point is
+    # clicked using this widget
+    # http://stackoverflow.com/questions/4652439/
+    # is-there-a-matplotlib-equivalent-of-matlabs-datacursormode/
+    # 4674445#4674445
     if obs_labels is True:
         if data is not None:
             obs_labels = data.index

statsmodels/graphics/tests/test_regressionplots.py

@@ -140,9 +140,9 @@ class TestPlotFormula(TestPlotPandas):
     def test_one_column_exog(self, close_figures):
         from statsmodels.formula.api import ols
         res = ols("y~var1-1", data=self.data).fit()
-        fig = plot_regress_exog(res, "var1")
+        plot_regress_exog(res, "var1")
         res = ols("y~var1", data=self.data).fit()
-        fig = plot_regress_exog(res, "var1")
+        plot_regress_exog(res, "var1")
 
 
 class TestABLine(object):

statsmodels/sandbox/predict_functional.py

@@ -426,7 +426,7 @@ def _glm_basic_scr(result, exog, alpha):
     # Calculate kappa_0 (formula 42 from Sun et al)
     bz = np.linalg.solve(B.T, exog.T).T
     bz /= np.sqrt(n)
-    bz /= sigma[:, None]
+    bz /= np.asarray(sigma)[:, None]
     bzd = np.diff(bz, 1, axis=0)
     bzdn = (bzd**2).sum(1)
     kappa_0 = np.sqrt(bzdn).sum()

statsmodels/sandbox/tests/test_predict_functional.py

@@ -63,6 +63,8 @@ def test_formula(self, close_figures):
         plt.clf()
         fig = plt.figure()
         ax = plt.axes([0.1, 0.1, 0.7, 0.8])
+        fvals1, pr1 = np.asarray(fvals1), np.asarray(pr1)
+        fvals2, pr2 = np.asarray(fvals2), np.asarray(pr2)
         plt.plot(fvals1, pr1, '-', label='x4=B')
         plt.plot(fvals2, pr2, '-', label='x4=C')
         ha, lb = ax.get_legend_handles_labels()
@@ -110,6 +112,7 @@ def test_lm_contrast(self, close_figures):
         plt.clf()
         fig = plt.figure()
         ax = plt.axes([0.1, 0.1, 0.67, 0.8])
+        fvals, pr = np.asarray(fvals), np.asarray(pr)
         plt.plot(fvals, pr, '-', label="Estimate", color='orange', lw=4)
         plt.plot(fvals, 4 - fvals, '-', label="Truth", color='lime', lw=4)
         plt.fill_between(fvals, cb[:, 0], cb[:, 1], color='grey')
@@ -146,6 +149,7 @@ def test_glm_formula_contrast(self, close_figures):
         plt.clf()
         fig = plt.figure()
         ax = plt.axes([0.1, 0.1, 0.67, 0.8])
+        fvals, pr = np.asarray(fvals), np.asarray(pr)
         plt.plot(fvals, pr, '-', label="Estimate", color='orange', lw=4)
         plt.plot(fvals, 0.2 - 0.1*fvals, '-', label="Truth", color='lime', lw=4)
         plt.fill_between(fvals, cb[:, 0], cb[:, 1], color='grey')
@@ -259,6 +263,9 @@ def test_glm_formula(self, close_figures):
             plt.clf()
             fig = plt.figure()
             ax = plt.axes([0.1, 0.1, 0.7, 0.8])
+            fvals1, pr1 = np.asarray(fvals1), np.asarray(pr1)
+            fvals2, pr2 = np.asarray(fvals2), np.asarray(pr2)
+            exact1, exact2 = np.asarray(exact1), np.asarray(exact2)
             plt.plot(fvals1, pr1, '-', label='x3=B')
             plt.plot(fvals2, pr2, '-', label='x3=C')
             plt.plot(fvals1, exact1, '-', label='x3=B (exact)')

statsmodels/tsa/ar_model.py

@@ -2093,7 +2093,7 @@ def plot_diagnostics(self, lags=10, fig=None, figsize=None):
             hold_back = self.model.hold_back
             x = hold_back + np.arange(self.resid.shape[0])
         std_resid = resid / np.sqrt(self.sigma2)
-        ax.plot(x, std_resid)
+        ax.plot(np.asarray(x), np.asarray(std_resid))
         ax.hlines(0, x[0], x[-1], alpha=0.5)
         ax.set_xlim(x[0], x[-1])
         ax.set_title('Standardized residual')

statsmodels/tsa/holtwinters.py

@@ -680,12 +680,13 @@ def fit(self, smoothing_level=None, smoothing_slope=None, smoothing_seasonal=Non
                 # using guesstimates for the levels
                 txi = xi & np.array([True, True, True, False, False, True] + [False] * m)
                 txi = txi.astype(np.bool)
-                bounds = np.array([(0.0, 1.0), (0.0, 1.0), (0.0, 1.0),
-                                   (0.0, None), (0.0, None), (0.0, 1.0)] + [(None, None), ] * m)
+                bounds = ([(0.0, 1.0), (0.0, 1.0), (0.0, 1.0), (0.0, None),
+                           (0.0, None), (0.0, 1.0)] + [(None, None), ] * m)
                 args = (txi.astype(np.uint8), p, y, lvls, b, s, m, self.nobs,
                         max_seen)
                 if start_params is None and np.any(txi) and use_brute:
-                    res = brute(func, bounds[txi], args, Ns=20,
+                    _bounds = [b for b, flag in zip(bounds, txi) if flag]
+                    res = brute(func, _bounds, args, Ns=20,
                                 full_output=True, finish=None)
                     p[txi], max_seen, _, _ = res
                 else:
@@ -708,14 +709,22 @@ def fit(self, smoothing_level=None, smoothing_slope=None, smoothing_seasonal=Non
                     # Take a deeper look in the local minimum we are in to find the best
                     # solution to parameters, maybe hop around to try escape the local
                     # minimum we may be in.
+                    _bounds = [b for b, flag in zip(bounds, xi) if flag]
                     res = basinhopping(func, p[xi],
-                                       minimizer_kwargs={'args': args, 'bounds': bounds[xi]},
+                                       minimizer_kwargs={'args': args, 'bounds': _bounds},
                                        stepsize=0.01)
                     success = res.lowest_optimization_result.success
                 else:
                     # Take a deeper look in the local minimum we are in to find the best
                     # solution to parameters
-                    res = minimize(func, p[xi], args=args, bounds=bounds[xi])
+                    _bounds = [b for b, flag in zip(bounds, xi) if flag]
+                    lb, ub = np.asarray(_bounds).T.astype(np.float)
+                    initial_p = p[xi]
+                    loc = p[xi] < lb
+                    initial_p[loc] = lb[loc]
+                    loc = p[xi] > ub
+                    initial_p[loc] = ub[loc]
+                    res = minimize(func, p[xi], args=args, bounds=_bounds)
                     success = res.success
 
                 if not success:

statsmodels/tsa/tests/test_holtwinters.py

@@ -323,8 +323,6 @@ def test_holt_damp_R(self):
             447.2614880558126]
         assert_allclose(fit.forecast(10), desired, atol=1e-4)
 
-
-
     def test_hw_seasonal(self):
         fit1 = ExponentialSmoothing(self.aust, seasonal_periods=4,
                                     trend='additive',