v0.2.8

.gitignore

@@ -1,5 +1,6 @@
 # Byte-compiled / optimized / DLL files
 __pycache__/
+.idea/*
 *.py[cod]
 *$py.class
 

dabest/__init__.py

@@ -23,4 +23,4 @@
 from ._stats_tools import effsize as effsize
 from ._classes import TwoGroupsEffectSize 
 
-__version__ = "0.2.7"
+__version__ = "0.2.8"

dabest/_classes.py

@@ -449,6 +449,9 @@ def __init__(self, control, test, effect_size,
          'pct_low': -0.763588353717278,
          'pvalue_brunner_munzel': nan,
          'pvalue_kruskal': nan,
+         'pvalue_lqrt_paired': nan,
+         'pvalue_lqrt_unpaired_equal_variance': 0.36,
+         'pvalue_lqrt_unpaired_unequal_variance': 0.36,
          'pvalue_mann_whitney': 0.2600723060808019,
          'pvalue_paired_students_t': nan,
          'pvalue_students_t': 0.34743913903372836,
@@ -458,6 +461,9 @@ def __init__(self, control, test, effect_size,
          'resamples': 5000,
          'statistic_brunner_munzel': nan,
          'statistic_kruskal': nan,
+         'statistic_lqrt_paired': nan,
+         'statistic_lqrt_unpaired_equal_variance': 0.8894980773231964,
+         'statistic_lqrt_unpaired_unequal_variance': 0.8916901409507432,
          'statistic_mann_whitney': 406.0,
          'statistic_paired_students_t': nan,
          'statistic_students_t': 0.9472545159069105,
@@ -470,6 +476,7 @@ def __init__(self, control, test, effect_size,
         from numpy.random import choice, seed
 
         import scipy.stats as spstats
+        import lqrt
 
         # import statsmodels.stats.power as power
 
@@ -601,6 +608,12 @@ def __init__(self, control, test, effect_size,
             wilcoxon = spstats.wilcoxon(control, test)
             self.__pvalue_wilcoxon = wilcoxon.pvalue
             self.__statistic_wilcoxon = wilcoxon.statistic
+
+            lqrt_result = lqrt.lqrtest_rel(control, test, 
+                                    random_state=random_seed)
+
+            self.__pvalue_paired_lqrt = lqrt_result.pvalue
+            self.__statistic_paired_lqrt = lqrt_result.statistic
 
             if effect_size != "median_diff":
                 # Paired Student's t-test.
@@ -658,7 +671,24 @@ def __init__(self, control, test, effect_size,
                 # in terms of rank (eg. all zeros.)
                 pass
 
+            # Likelihood Q-Ratio test:
+            lqrt_equal_var_result = lqrt.lqrtest_ind(control, test, 
+                                        random_state=random_seed,
+                                        equal_var=True)
+
+            self.__pvalue_lqrt_equal_var = lqrt_equal_var_result.pvalue
+            self.__statistic_lqrt_equal_var = lqrt_equal_var_result.statistic
+
+            lqrt_unequal_var_result = lqrt.lqrtest_ind(control, test, 
+                                        random_state=random_seed,
+                                        equal_var=False)
+
+            self.__pvalue_lqrt_unequal_var = lqrt_unequal_var_result.pvalue
+            self.__statistic_lqrt_unequal_var = lqrt_unequal_var_result.statistic
+
+
             standardized_es = es.cohens_d(control, test, is_paired=False)
+
             # self.__power = power.tt_ind_solve_power(standardized_es,
             #                                         len(control),
             #                                         alpha=self.__alpha,
@@ -968,6 +998,65 @@ def statistic_mann_whitney(self):
 
 
 
+
+    @property
+    def pvalue_lqrt_paired(self):
+        from numpy import nan as npnan
+        try:
+            return self.__pvalue_paired_lqrt
+        except AttributeError:
+            return npnan
+
+
+
+    @property
+    def statistic_lqrt_paired(self):
+        from numpy import nan as npnan
+        try:
+            return self.__statistic_paired_lqrt
+        except AttributeError:
+            return npnan
+
+
+    @property
+    def pvalue_lqrt_unpaired_equal_variance(self):
+        from numpy import nan as npnan
+        try:
+            return self.__pvalue_lqrt_equal_var
+        except AttributeError:
+            return npnan
+
+
+
+    @property
+    def statistic_lqrt_unpaired_equal_variance(self):
+        from numpy import nan as npnan
+        try:
+            return self.__statistic_lqrt_equal_var
+        except AttributeError:
+            return npnan
+
+
+    @property
+    def pvalue_lqrt_unpaired_unequal_variance(self):
+        from numpy import nan as npnan
+        try:
+            return self.__pvalue_lqrt_unequal_var
+        except AttributeError:
+            return npnan
+
+
+
+    @property
+    def statistic_lqrt_unpaired_unequal_variance(self):
+        from numpy import nan as npnan
+        try:
+            return self.__statistic_lqrt_unequal_var
+        except AttributeError:
+            return npnan
+
+
+
     # @property
     # def power(self):
     #     from numpy import nan as npnan
@@ -1089,7 +1178,16 @@ def __pre_calc(self):
                             'statistic_paired_students_t',
 
                             'pvalue_kruskal',
-                            'statistic_kruskal']
+                            'statistic_kruskal',
+
+                            'pvalue_lqrt_paired',
+                            'statistic_lqrt_paired', 
+
+                            'pvalue_lqrt_unpaired_equal_variance', 
+                            'statistic_lqrt_unpaired_equal_variance',
+
+                            'pvalue_lqrt_unpaired_unequal_variance',
+                            'statistic_lqrt_unpaired_unequal_variance']
 
         self.__results   = out_.reindex(columns=columns_in_order)
         self.__results.dropna(axis="columns", how="all", inplace=True)
@@ -1209,6 +1307,12 @@ def plot(self, color_col=None,
             pyplot.violinplot` command here, as a dict. If None, the following
             keywords are passed to violinplot : {'widths':0.5, 'vert':True,
             'showextrema':False, 'showmedians':False}.
+        slopegraph_kwargs : dict, default None
+            This will change the appearance of the lines used to join each pair
+            of observations when `show_pairs=True`. Pass any keyword arguments
+            accepted by matplotlib `plot()` function here, as a dict.
+            If None, the following keywords are
+            passed to plot() : {'linewidth':1, 'alpha':0.5}.
         reflines_kwargs : dict, default None
             This will change the appearance of the zero reference lines. Pass
             any keyword arguments accepted by the matplotlib Axes `hlines`

dabest/plotter.py

@@ -130,7 +130,7 @@ def EffectSizeDataFramePlotter(EffectSizeDataFrame, **plot_kwargs):
     if plot_kwargs["slopegraph_kwargs"] is None:
         slopegraph_kwargs = default_slopegraph_kwargs
     else:
-        slopegraph_kwargs = merge_two_dicts(slopegraph_kwargs,
+        slopegraph_kwargs = merge_two_dicts(default_slopegraph_kwargs,
                                             plot_kwargs["slopegraph_kwargs"])
 
 
@@ -697,7 +697,7 @@ def EffectSizeDataFramePlotter(EffectSizeDataFrame, **plot_kwargs):
         else:
             contrast_ylim_high, contrast_ylim_low = contrast_axes_ylim
         if contrast_ylim_low < 0 < contrast_ylim_high:
-            contrast_axes.axhline(y=0, lw=0.75, color=ytick_color)
+            contrast_axes.axhline(y=0, **reflines_kwargs)
 
         # Compute the end of each x-axes line.
         rightend_ticks = np.array([len(i)-1 for i in idx]) + np.array(ticks_to_skip)

dabest/tests/test_01_effsizes_pvals.py

@@ -5,9 +5,10 @@
 
 
 import sys
+import pytest
+import lqrt
 import numpy as np
 import scipy as sp
-import pytest
 import pandas as pd
 from .._stats_tools import effsize
 from .._classes import TwoGroupsEffectSize
@@ -176,4 +177,31 @@ def test_ordinal_dominance():
                              "cliffs_delta", is_paired=False)
 
     p1 = sp.stats.brunnermunzel(likert_control, likert_treatment).pvalue
-    assert es.pvalue_brunner_munzel == pytest.approx(p1)
+    assert es.pvalue_brunner_munzel == pytest.approx(p1)
+
+
+
+def test_lqrt_unpaired():
+    es = TwoGroupsEffectSize(wellbeing.control, wellbeing.expt, 
+                             "mean_diff", is_paired=False)
+
+    p1 = lqrt.lqrtest_ind(wellbeing.control, wellbeing.expt,
+                          equal_var=False,
+                          random_state=12345)
+
+    p2 = lqrt.lqrtest_ind(wellbeing.control, wellbeing.expt,
+                          equal_var=True,
+                          random_state=12345)
+
+    assert es.pvalue_lqrt_unpaired_unequal_variance == pytest.approx(p1.pvalue)
+    assert es.pvalue_lqrt_unpaired_equal_variance == pytest.approx(p2.pvalue)
+
+
+def test_lqrt_paired():
+    es = TwoGroupsEffectSize(paired_wellbeing.pre, paired_wellbeing.post, 
+                             "mean_diff", is_paired=True)
+
+    p1 = lqrt.lqrtest_rel(paired_wellbeing.pre, paired_wellbeing.post, 
+                 random_state=12345)
+
+    assert es.pvalue_lqrt_paired == pytest.approx(p1.pvalue)

dabest/tests/test_03_plotting.py

@@ -178,7 +178,7 @@ def test_14_gardner_altman_paired_color():
 
 @pytest.mark.mpl_image_compare
 def test_15_change_palette_a():
-    return multi_2group.mean_diff.plot(fig_size=(7, 6),
+    return multi_2group.mean_diff.plot(fig_size=(8, 6),
                                        color_col="Gender", 
                                        custom_palette="Dark2");
 
@@ -285,7 +285,7 @@ def test_23_ticker_cumming():
 
 
 @pytest.mark.mpl_image_compare
-def test_24_test_wide_df_nan():
+def test_24_wide_df_nan():
 
     wide_df_dabest = load(wide_df, 
                           idx=("Control", "Test 1", "Test 2", "Test 3")
@@ -296,16 +296,59 @@ def test_24_test_wide_df_nan():
 
 
 @pytest.mark.mpl_image_compare
-def test_25_test_long_df_nan():
+def test_25_long_df_nan():
 
     long_df_dabest = load(long_df, x="group", y="value",
                           idx=("Control", "Test 1", "Test 2", "Test 3")
                           )
 
     return long_df_dabest.mean_diff.plot(); 
+
+
+
+@pytest.mark.mpl_image_compare
+def test_26_slopegraph_kwargs():
+
+    return two_groups_paired.mean_diff.plot(
+            slopegraph_kwargs=dict(linestyle='dotted')
+            );
+
+
+
+@pytest.mark.mpl_image_compare
+def test_27_gardner_altman_reflines_kwargs():
+
+    return two_groups_unpaired.mean_diff.plot(
+            reflines_kwargs=dict(linestyle='dotted')
+            );
 
 
 
+@pytest.mark.mpl_image_compare
+def test_28_unpaired_cumming_reflines_kwargs():
+
+    return two_groups_unpaired.mean_diff.plot(
+            fig_size=(12,10),
+            float_contrast=False,
+            reflines_kwargs=dict(linestyle='dotted', 
+                                 linewidth=2),
+            contrast_ylim=(-1, 1)
+            );
+
+
+
+@pytest.mark.mpl_image_compare
+def test_28_paired_cumming_slopegraph_reflines_kwargs():
+
+    return two_groups_paired.mean_diff.plot(float_contrast=False,
+                                 color_col="Gender",
+                                 slopegraph_kwargs=dict(linestyle='dotted'),
+                                 reflines_kwargs=dict(linestyle='dashed', 
+                                                      linewidth=2),
+                                 contrast_ylim=(-1, 1)
+                                 );
+
+
 @pytest.mark.mpl_image_compare
 def test_99_style_sheets():
     # Perform this test last so we don't have to reset the plot style.

docs/source/about.rst

@@ -14,12 +14,17 @@ To find out more about the authors’ research, please visit the `Claridge-Chang
 Contributors
 ------------
 
+- DizietAsahi (`DizietAsahi <https://github.com/DizietAsahi>`_) with `PR #86 <https://github.com/ACCLAB/DABEST-python/pull/86>`_:  Fix bugs in slopegraph and reference line keyword parsing.
+
+- Adam Li (`@adam2392 <https://github.com/adam2392>`_) with `PR #85 <https://github.com/ACCLAB/DABEST-python/pull/85>`_:  Implement `Lq-Likelihood-Ratio-Type Test <https://github.com/alyakin314/lqrt>`_ in statistical output.
+
 - Mason Malone (`@MasonM <https://github.com/MasonM>`_) with `PR #30 <https://github.com/ACCLAB/DABEST-python/pull/30>`_:  Fix plot error when effect size is 0.
 
 - Matthew Edwards (`@mje-nz <https://github.com/mje-nz>`_) with `PR #71 <https://github.com/ACCLAB/DABEST-python/pull/30>`_: Specify dependencies correctly in ``setup.py``. 
 
 - Adam Nekimken (`@anekimken <https://github.com/anekimken>`_) with `PR #73 <https://github.com/ACCLAB/DABEST-python/pull/73>`_: Implement inset axes so estimation plots can be plotted on a pre-determined :py:mod:`matplotlib` :py:class:`Axes` object.
 
+
 Typography
 ----------
 
@@ -31,7 +36,7 @@ License
 
 The DABEST package in Python is licenced under the `BSD 3-clause Clear License <https://choosealicense.com/licenses/bsd-3-clause-clear/>`_.
 
-Copyright (c) 2016-2019, Joses W. Ho
+Copyright (c) 2016-2020, Joses W. Ho
 All rights reserved.
 
 Redistribution and use in source and binary forms, with or without

docs/source/conf.py

@@ -77,7 +77,7 @@
 # The short X.Y version.
 version = '0.2'
 # The full version, including alpha/beta/rc tags.
-release = '0.2.7'
+release = '0.2.8'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.

docs/source/getting-started.rst

@@ -10,13 +10,14 @@ Requirements
 
 Python 3.7 is strongly recommended. DABEST has also been tested with Python 3.5 and 3.6.
 
-In addition, the following packages are also required:
+In addition, the following packages are also required (listed with their minimal versions):
 
 * `numpy 1.15 <https://www.numpy.org>`_
 * `scipy 1.2 <https://www.scipy.org>`_
 * `matplotlib 3.0 <https://www.matplotlib.org>`_
-* `pandas 0.24 <https://pandas.pydata.org>`_
+* `pandas 0.25.3 <https://pandas.pydata.org>`_
 * `seaborn 0.9 <https://seaborn.pydata.org>`_
+* `lqrt 0.3.2 <https://github.com/alyakin314/lqrt>`_
 
 To obtain these package dependencies easily, it is highly recommended to download the `Anaconda <https://www.continuum.io/downloads>`_ distribution of Python.
 

docs/source/index.rst

@@ -18,8 +18,11 @@ Analyze your data with estimation statistics!
 
 News
 ----
+December 2019:
+  - v0.2.8 released. This release adds the `Lq-Likelihood-Ratio-Type Test <https://github.com/alyakin314/lqrt>`_ in the statistical output, and also a bugfix for  slopegraph and reference line keyword parsing. For more information, see the :doc:`release-notes`.
+
 October 2019:
-  - v0.2.7 released. A minor bugfix in the handling of wide datasets with unequal Ns in each group. For more information, see the :doc:`release-notes`.
+  - v0.2.7 released. A minor bugfix in the handling of wide datasets with unequal Ns in each group. 
   - v0.2.6 released. This release has one new feature (plotting of estimation plot inside any :py:mod:`matplotlib` :py:class:`Axes`; see the section on :ref:`inset_plot` in the :doc:`tutorial`). There are also two bug patches for the handling of bootstrap plotting, and of dependency installation. 
 
 September 2019: