Merge pull request #5 from phobson/y-oriented-tests

tests for when Y is the prob axis

Author: phobson

.travis.yml

@@ -34,9 +34,7 @@ install:
   - conda create --yes -n test python=$TRAVIS_PYTHON_VERSION
   - source activate test
   #- conda config --add channels phobson
-  - conda install --yes numpy nose scipy
-  - pip install matplotlib
-  - conda install --yes coverage docopt requests pyyaml
+  - conda install --yes numpy nose scipy matplotlib=1.4 coverage docopt requests pyyaml
   - pip install coveralls
   - pip install .
 

probscale/probscale.py

@@ -88,25 +88,25 @@ def _sig_figs(cls, x, n, expthresh=5, forceint=False):
 
         """
 
+        # return a string value unaltered
+        if isinstance(x, str) or x == 0.0:
+            out = str(x)
+
         # check on the number provided
-        if x is not None and not np.isinf(x) and not np.isnan(x):
+        elif x is not None and not np.isinf(x) and not np.isnan(x):
 
             # check on the _sig_figs
             if n < 1:
                 raise ValueError("number of sig figs (n) must be greater than zero")
 
-            # return a string value unaltered
-            if isinstance(x, str):
-                out = x
-
             elif forceint:
                 out = '{:,.0f}'.format(x)
 
             # logic to do all of the rounding
-            elif x != 0.0:
+            else:
                 order = np.floor(np.log10(np.abs(x)))
 
-                if -1.0 * expthresh <= order <= expthresh:
+                if (-1.0 * expthresh <= order <= expthresh):
                     decimal_places = int(n - 1 - order)
 
                     if decimal_places <= 0:
@@ -121,9 +121,6 @@ def _sig_figs(cls, x, n, expthresh=5, forceint=False):
                     fmt = '{0:.%de}' % decimal_places
                     out = fmt.format(x)
 
-            else:
-                out = str(round(x, n))
-
         # with NAs and INFs, just return 'NA'
         else:
             out = 'NA'

probscale/tests/test_probscale/baseline_images/test_probscale/test_viz/test_probplot_pp_bestfit_probax_y.png

@@ -81,6 +81,16 @@ def test_baseline(self):
         nt.assert_equal(self.fmt._sig_figs(self.x, 3), self.known_3)
         nt.assert_equal(self.fmt._sig_figs(self.x, 4), self.known_4)
 
+    def test_string(self):
+        nt.assert_equal(self.fmt._sig_figs('1.23', 3), '1.23')
+
+    def test_na_inf(self):
+        nt.assert_equal(self.fmt._sig_figs(np.nan, 3), 'NA')
+        nt.assert_equal(self.fmt._sig_figs(np.inf, 3), 'NA')
+
+    def test_zero(self):
+        nt.assert_equal(self.fmt._sig_figs(0, 3), '0')
+
     def test_trailing_zeros(self):
         nt.assert_equal(self.fmt._sig_figs(self.x, 8), self.known_8)
 

probscale/tests/test_probscale/baseline_images/test_probscale/test_viz/test_probplot_pp_probax_y.png

@@ -1,4 +1,4 @@
-import matplotlib
+import matplotlib
 matplotlib.use('agg')
 
 import numpy as np
@@ -24,12 +24,12 @@ def setup_plot_data():
     return data
 
 
-@cleanup
 class Test__check_ax_obj(object):
     @nt.raises(ValueError)
     def test_bad_value(self):
         viz._check_ax_obj('junk')
 
+    @cleanup
     def test_with_ax(self):
         fig, ax = plt.subplots()
         fig1, ax1 = viz._check_ax_obj(ax)
@@ -38,15 +38,16 @@ def test_with_ax(self):
         nt.assert_true(ax1 is ax)
         nt.assert_true(fig1 is fig)
 
+    @cleanup
     def test_with_None(self):
         fig1, ax1 = viz._check_ax_obj(None)
         nt.assert_true(isinstance(ax1, plt.Axes))
         nt.assert_true(isinstance(fig1, plt.Figure))
 
 
-class Test__check_fig_arg(object):
+class Test__check_fit_arg(object):
     @nt.raises(ValueError)
-    def test_bad_value(self):
+    def test_bad_fitarg(self):
         viz._check_fit_arg('junk', 'fitprobs')
 
     def test_x(self):
@@ -67,8 +68,8 @@ def test_None(self):
 
 
 class Test__check_ax_name(object):
-    @nt.raises(ValueError)
-    def test_bad_value(self):
+    @nt.raises
+    def test_bad_name(self):
         viz._check_fit_arg('junk', 'axname')
 
     def test_x(self):
@@ -270,6 +271,17 @@ def test_linlog(self):
         )
 
 
+class Test__check_ax_type(object):
+    @nt.raises(ValueError)
+    def test_bad_value(self):
+        viz._check_ax_type("JUNK")
+
+    def test_upper(self):
+        nt.assert_equal('pp', viz._check_ax_type('PP'))
+        nt.assert_equal('qq', viz._check_ax_type('QQ'))
+        nt.assert_equal('prob', viz._check_ax_type('ProB'))
+
+
 @image_comparison(baseline_images=['test_probplot_prob'], extensions=['png'])
 def test_probplot_prob():
     fig, ax = plt.subplots()
@@ -319,3 +331,65 @@ def test_probplot_pp_bestfit():
     fig = viz.probplot(data, ax=ax, axtype='pp', otherscale='linear',
                        xlabel='test x', bestfit=True, ylabel='test y',
                        scatter_kws=scatter_kws, line_kws=line_kws)
+
+
+@image_comparison(baseline_images=['test_probplot_prob_probax_y'], extensions=['png'])
+def test_probplot_prob_probax_y():
+    fig, ax = plt.subplots()
+    data = setup_plot_data()
+    fig = viz.probplot(data, ax=ax, xlabel='Test xlabel', otherscale='log', probax='y')
+    nt.assert_true(isinstance(fig, plt.Figure))
+
+
+@image_comparison(baseline_images=['test_probplot_qq_probax_y'], extensions=['png'])
+def test_probplot_qq_probax_y():
+    fig, ax = plt.subplots()
+    data = setup_plot_data()
+    fig = viz.probplot(data, ax=ax, axtype='qq', ylabel='Test label', probax='y',
+                       otherscale='log', scatter_kws=dict(color='r'))
+
+
+@image_comparison(baseline_images=['test_probplot_pp_probax_y'], extensions=['png'])
+def test_probplot_pp_probax_y():
+    fig, ax = plt.subplots()
+    data = setup_plot_data()
+    scatter_kws = dict(color='b', linestyle='--', markeredgecolor='g', markerfacecolor='none')
+    fig = viz.probplot(data, ax=ax, axtype='pp', otherscale='linear', probax='y',
+                       xlabel='test x', ylabel='test y', scatter_kws=scatter_kws)
+
+
+@image_comparison(baseline_images=['test_probplot_prob_bestfit_probax_y'], extensions=['png'])
+def test_probplot_prob_bestfit_probax_y():
+    fig, ax = plt.subplots()
+    data = setup_plot_data()
+    fig = viz.probplot(data, ax=ax, xlabel='Test xlabel', bestfit=True, otherscale='log', probax='y')
+    nt.assert_true(isinstance(fig, plt.Figure))
+
+
+@image_comparison(baseline_images=['test_probplot_qq_bestfit_probax_y'], extensions=['png'])
+def test_probplot_qq_bestfit_probax_y():
+    fig, ax = plt.subplots()
+    data = setup_plot_data()
+    fig = viz.probplot(data, ax=ax, axtype='qq', bestfit=True, ylabel='Test label', otherscale='log', probax='y')
+
+
+@image_comparison(baseline_images=['test_probplot_pp_bestfit_probax_y'], extensions=['png'])
+def test_probplot_pp_bestfit_probax_y():
+    fig, ax = plt.subplots()
+    data = setup_plot_data()
+    scatter_kws = {'marker': 's', 'color': 'red'}
+    line_kws = {'linestyle': '--', 'linewidth': 3}
+    fig = viz.probplot(data, ax=ax, axtype='pp', otherscale='linear', probax='y',
+                       xlabel='test x', bestfit=True, ylabel='test y',
+                       scatter_kws=scatter_kws, line_kws=line_kws)
+
+
+@cleanup
+def test_probplot_test_results():
+    fig, ax = plt.subplots()
+    data = setup_plot_data()
+    fig, results = viz.probplot(data, return_results=True)
+
+    nt.assert_true(isinstance(results, dict))
+    known_keys = sorted(['q', 'x', 'y', 'xhat', 'yhat', 'res'])
+    nt.assert_list_equal(sorted(list(results.keys())), known_keys)