v0.2.6

.travis.yml

@@ -22,8 +22,7 @@ install:
   - conda config --add channels conda-forge
   - conda create -n testenv --yes pip python=$PYTHON matplotlib
   - source activate testenv
-  - pip install pytest==4.3
-  - pip install .
+  - pip install .[dev]
 
 
 script: pytest dabest

CONTRIBUTING.md

@@ -8,9 +8,9 @@
 
 
 ## Did you write a patch that fixes a bug?
-- Open a new GitHub [pull request](https://help.github.com/en/articles/about-pull-requests)(PR for short) with the patch.
+- Open a new GitHub [pull request](https://help.github.com/en/articles/about-pull-requests) (PR for short) with the patch.
 
-- Create the PR into the development branch, which is indicated by `v{latest version number}_dev`.
+- Create the PR into the development branch, which is indicated by `v{latest version number}-dev`.
 
 - Clearly state the problem and solution in the PR description. Include the relevant [issue number](https://guides.github.com/features/issues/) if applicable.
 

README.md

@@ -6,6 +6,7 @@
 [![Free-to-view citation](https://zenodo.org/badge/DOI/10.1038/s41592-019-0470-3.svg)](https://rdcu.be/bHhJ4)
 [![License](https://img.shields.io/badge/License-BSD%203--Clause--Clear-orange.svg)](https://spdx.org/licenses/BSD-3-Clause-Clear.html)
 
+
 ## About
 
 DABEST is a package for **D**ata **A**nalysis using **B**ootstrap-Coupled **EST**imation.
@@ -22,21 +23,12 @@ An estimation plot has two key features.
 
 DABEST powers [estimationstats.com](https://www.estimationstats.com/), allowing everyone access to high-quality estimation plots.
 
-## Requirements
-
-DABEST has been tested on Python 3.5, 3.6, and 3.7.
-
-In addition, the following packages are also required:
-- [numpy](https://www.numpy.org) (1.15)
-- [scipy](https://www.scipy.org) (1.2)
-- [matplotlib](https://www.matplotlib.org) (3.0)
-- [seaborn](https://seaborn.pydata.org) (0.9)
-- [pandas](https://pandas.pydata.org) (0.24).
-
-To obtain these package dependencies easily, it is highly recommended to download the [Anaconda distribution](https://www.continuum.io/downloads) of Python.
 
 ## Installation
 
+This package is tested on Python 3.5, 3.6, and 3.7.
+It is highly recommended to download the [Anaconda distribution](https://www.continuum.io/downloads) of Python in order to obtain the dependencies easily.
+
 You can install this package via `pip`.
 
 To install, at the command line run
@@ -56,6 +48,7 @@ Then, navigate to the cloned repo in the command line and run
 pip install .
 ```
 
+
 ## Usage
 
 ```python3
@@ -99,6 +92,7 @@ All contributions are welcome; please read the [Guidelines for contributing](htt
 
 We also have a [Code of Conduct](https://github.com/ACCLAB/DABEST-python/blob/master/CODE_OF_CONDUCT.md) to foster an inclusive and productive space.
 
+
 ## Acknowledgements
 
 We would like to thank alpha testers from the [Claridge-Chang lab](https://www.claridgechang.net/): [Sangyu Xu](https://github.com/sangyu), [Xianyuan Zhang](https://github.com/XYZfar), [Farhan Mohammad](https://github.com/farhan8igib), Jurga Mituzaitė, and Stanislav Ott.

dabest/__init__.py

@@ -23,4 +23,4 @@
 from ._stats_tools import effsize as effsize
 from ._classes import TwoGroupsEffectSize 
 
-__version__ = "0.2.5"
+__version__ = "0.2.6"

dabest/_classes.py

@@ -460,7 +460,7 @@ def __init__(self, control, test, effect_size,
          'statistic_wilcoxon': nan}
         """
 
-        from numpy import array, isnan
+        from numpy import array, isnan, isinf
         from numpy import sort as npsort
         from numpy.random import choice, seed
 
@@ -522,6 +522,20 @@ def __init__(self, control, test, effect_size,
                             control, test, is_paired, effect_size,
                             resamples, random_seed)
         self.__bootstraps = npsort(bootstraps)
+
+        # Added in v0.2.6.
+        # Raises a UserWarning if there are any infiinities in the bootstraps.
+        num_infinities = len(self.__bootstraps[isinf(self.__bootstraps)])
+
+        if num_infinities > 0:
+            warn_msg = "There are {} bootstrap(s) that are not defined. "\
+            "This is likely due to smaple sample sizes. "\
+            "The values in a bootstrap for a group will be more likely "\
+            "to be all equal, with a resulting variance of zero. "\
+            "The computation of Cohen's d and Hedges' g thus "\
+            "involved a division by zero. "
+            warnings.warn(warn_msg.format(num_infinities), 
+                          category=UserWarning)
 
         self.__bias_correction = ci2g.compute_meandiff_bias_correction(
                                     self.__bootstraps, self.__difference)
@@ -1103,6 +1117,7 @@ def plot(self, color_col=None,
 
             fig_size=None,
             dpi=100,
+            ax=None,
 
             swarmplot_kwargs=None,
             violinplot_kwargs=None,
@@ -1112,6 +1127,7 @@ def plot(self, color_col=None,
             legend_kwargs=None):
         """
         Creates an estimation plot for the effect size of interest.
+
 
         Parameters
         ----------
@@ -1176,6 +1192,9 @@ def plot(self, color_col=None,
             The desired dimensions of the figure as a (length, width) tuple.
         dpi : int, default 100
             The dots per inch of the resulting figure.
+        ax : matplotlib.Axes, default None
+            Provide an existing Axes for the plots to be created. If no Axes is
+            specified, a new matplotlib Figure will be created.
         swarmplot_kwargs : dict, default None
             Pass any keyword arguments accepted by the seaborn `swarmplot`
             command here, as a dict. If None, the following keywords are
@@ -1206,9 +1225,14 @@ def plot(self, color_col=None,
 
         Returns
         -------
-        A :class:`matplotlib.figure.Figure` with 2 Axes.
-
+        A :class:`matplotlib.figure.Figure` with 2 Axes, if ``ax = None``.
+        
         The first axes (accessible with ``FigName.axes[0]``) contains the rawdata swarmplot; the second axes (accessible with ``FigName.axes[1]``) has the bootstrap distributions and effect sizes (with confidence intervals) plotted on it.
+
+        If ``ax`` is specified, the rawdata swarmplot is accessed at ``ax`` 
+        itself, while the effect size axes is accessed at ``ax.contrast_axes``.
+        See the last example below.
+
 
         Examples
         --------
@@ -1244,6 +1268,14 @@ def plot(self, color_col=None,
         ...                                          "Test 2", "Test 3")
         ...                                 )
         >>> fig6 = my_shared_control.mean_diff.plot()
+
+        Creating estimation plots in individual panels of a figure.
+
+        >>> f, axx = plt.subplots(nrows=2, ncols=2, figsize=(15, 15))
+        >>> my_data.mean_diff.plot(ax=axx.flat[0])
+        >>> my_data_paired.mean_diff.plot(ax=axx.flat[1])
+        >>> my_shared_control.mean_diff.plot(ax=axx.flat[2])
+        >>> my_shared_control.mean_diff.plot(ax=axx.flat[3], float_contrast=False)
 
         """
 
@@ -1344,4 +1376,4 @@ def dabest_obj(self):
         Returns the `dabest` object that invoked the current EffectSizeDataFrame
         class.
         """
-        return self.__dabest_obj
+        return self.__dabest_obj

dabest/_stats_tools/confint_2group_diff.py

@@ -159,7 +159,24 @@ def compute_bootstrapped_diff(x0, x1, is_paired, effect_size,
 
     # reset seed
     np.random.seed()
-
+
+    # check whether there are any infinities in the bootstrap,
+    # which likely indicates the sample sizes are too small as
+    # the computation of Cohen's d and Hedges' g necessitated 
+    # a division by zero.
+    # Added in v0.2.6.
+
+    # num_infinities = len(out[np.isinf(out)])
+    # print(num_infinities)
+    # if num_infinities > 0:
+    #     warn_msg = "There are {} bootstraps that are not defined. "\
+    #     "This is likely due to smaple sample sizes. "\
+    #     "The values in a bootstrap for a group will be more likely "\
+    #     "to be all equal, with a resulting variance of zero. "\
+    #     "The computation of Cohen's d and Hedges' g will therefore "\
+    #     "involved a division by zero. "
+    #     warnings.warn(warn_msg.format(num_infinities), category="UserWarning")
+
     return out
 
 

dabest/_stats_tools/effsize.py

@@ -217,11 +217,13 @@ def cohens_d(control, test, is_paired=False):
         # assume the two arrays are ordered already.
         delta = test - control
         M = np.mean(delta)
-        return M / average_sd
+        divisor = average_sd
 
     else:
         M = np.mean(test) - np.mean(control)
-        return M / pooled_sd
+        divisor = pooled_sd
+
+    return M / divisor
 
 
 

dabest/plotter.py

@@ -31,6 +31,7 @@ def EffectSizeDataFramePlotter(EffectSizeDataFrame, **plot_kwargs):
 
         fig_size=None,
         dpi=100,
+        ax=None,
 
         swarmplot_kwargs=None,
         violinplot_kwargs=None,
@@ -254,28 +255,76 @@ def EffectSizeDataFramePlotter(EffectSizeDataFrame, **plot_kwargs):
 
     # Initialise the figure.
     # sns.set(context="talk", style='ticks')
-    init_fig_kwargs = dict(figsize=fig_size, dpi=plot_kwargs["dpi"])
+    init_fig_kwargs = dict(figsize=fig_size, dpi=plot_kwargs["dpi"],
+                           tight_layout=True)
+
+    width_ratios_ga = [2.5, 1]
+    h_space_cummings = 0.3
+    if plot_kwargs["ax"] is not None:
+        # New in v0.2.6.
+        # Use inset axes to create the estimation plot inside a single axes.
+        # Author: Adam L Nekimken. (PR #73)
+        inset_contrast = True
+        rawdata_axes = plot_kwargs["ax"]
+        ax_position = rawdata_axes.get_position()  # [[x0, y0], [x1, y1]]
+
+        fig = rawdata_axes.get_figure()
+
+        if float_contrast is True:
+            axins = rawdata_axes.inset_axes(
+                    [1, 0,
+                     width_ratios_ga[1]/width_ratios_ga[0], 1])
+            rawdata_axes.set_position(  # [l, b, w, h]
+                    [ax_position.x0,
+                     ax_position.y0,
+                     (ax_position.x1 - ax_position.x0) * (width_ratios_ga[0] /
+                                                         sum(width_ratios_ga)),
+                     (ax_position.y1 - ax_position.y0)])
+
+            contrast_axes = axins
 
-    # Here, we hardcode some figure parameters.
-    if float_contrast is True:
-        fig, axx = plt.subplots(ncols=2,
-                                gridspec_kw={"width_ratios": [2.5, 1],
-                                             "wspace": 0},
-                                **init_fig_kwargs)
+        else:
+            axins = rawdata_axes.inset_axes([0, -1 - h_space_cummings, 1, 1])
+            plot_height = ((ax_position.y1 - ax_position.y0) /
+                           (2 + h_space_cummings))
+            rawdata_axes.set_position(
+                    [ax_position.x0,
+                     ax_position.y0 + (1 + h_space_cummings) * plot_height,
+                     (ax_position.x1 - ax_position.x0),
+                     plot_height])
+
+            # If the contrast axes are NOT floating, create lists to store
+            # raw ylims and raw tick intervals, so that I can normalize
+            # their ylims later.
+            contrast_ax_ylim_low = list()
+            contrast_ax_ylim_high = list()
+            contrast_ax_ylim_tickintervals = list()
+        contrast_axes = axins
+        rawdata_axes.contrast_axes = axins
 
     else:
-        fig, axx = plt.subplots(nrows=2,
-                                gridspec_kw={"hspace": 0.3},
-                                **init_fig_kwargs)
-
-        # If the contrast axes are NOT floating, create lists to store raw ylims
-        # and raw tick intervals, so that I can normalize their ylims later.
-        contrast_ax_ylim_low = list()
-        contrast_ax_ylim_high = list()
-        contrast_ax_ylim_tickintervals = list()
+        inset_contrast = False
+        # Here, we hardcode some figure parameters.
+        if float_contrast is True:
+            fig, axx = plt.subplots(
+                    ncols=2,
+                    gridspec_kw={"width_ratios": width_ratios_ga,
+                                 "wspace": 0},
+                                 **init_fig_kwargs)
 
-    rawdata_axes  = axx[0]
-    contrast_axes = axx[1]
+        else:
+            fig, axx = plt.subplots(nrows=2,
+                                    gridspec_kw={"hspace": 0.3},
+                                    **init_fig_kwargs)
+            # If the contrast axes are NOT floating, create lists to store
+            # raw ylims and raw tick intervals, so that I can normalize
+            # their ylims later.
+            contrast_ax_ylim_low = list()
+            contrast_ax_ylim_high = list()
+            contrast_ax_ylim_tickintervals = list()
+
+        rawdata_axes  = axx[0]
+        contrast_axes = axx[1]
 
     rawdata_axes.set_frame_on(False)
     contrast_axes.set_frame_on(False)
@@ -423,7 +472,8 @@ def EffectSizeDataFramePlotter(EffectSizeDataFrame, **plot_kwargs):
         current_ci_high   = results.bca_high[j]
 
         # Create the violinplot.
-        v = contrast_axes.violinplot(current_bootstrap,
+        # New in v0.2.6: drop negative infinities before plotting.
+        v = contrast_axes.violinplot(current_bootstrap[~np.isinf(current_bootstrap)],
                                      positions=[tick],
                                      **violinplot_kwargs)
         # Turn the violinplot into half, and color it the same as the swarmplot.
@@ -651,19 +701,19 @@ def EffectSizeDataFramePlotter(EffectSizeDataFrame, **plot_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)
-
-        for ax in fig.axes:
+        
+        for ax in [rawdata_axes, contrast_axes]:
             sns.despine(ax=ax, bottom=True)
-
+        
             ylim = ax.get_ylim()
             xlim = ax.get_xlim()
             redraw_axes_kwargs['y'] = ylim[0]
-
+        
             for k, start_tick in enumerate(ticks_to_skip):
                 end_tick = rightend_ticks[k]
                 ax.hlines(xmin=start_tick, xmax=end_tick,
                           **redraw_axes_kwargs)
-
+        
             ax.set_ylim(ylim)
             del redraw_axes_kwargs['y']
 

dabest/pytest.ini

@@ -1,4 +1,9 @@
 [pytest]
 filterwarnings =
     ignore::UserWarning
-    ignore::DeprecationWarning
+    ignore::DeprecationWarning
+
+addopts = --mpl --mpl-baseline-path=dabest/tests/baseline_images
+
+markers =
+    mpl_image_compare: mark a test as implementing mpl image comparison.

dabest/tests/README.md

@@ -1,9 +1,8 @@
 # Testing
 
-We use [pytest](https://docs.pytest.org/en/latest) to execute the tests. More documentation of the testing paradigm will be added in the near future.
+We use [pytest](https://docs.pytest.org/en/latest) to execute the tests. For testing of plot generation, we use the [mpl plugin](https://github.com/matplotlib/pytest-mpl) for pytest. A range of different plots are created, and compared against the baseline images in the `baseline_images` subfolder.
 
 To run the tests, go to the root of this repo directory and run 
-
 ```shell
 pytest dabest
 ```