copulas 0.9.2.dev0__tar.gz → 0.10.0.dev0__tar.gz

{copulas-0.9.2.dev0 → copulas-0.10.0.dev0}/HISTORY.md

@@ -1,5 +1,18 @@
 # History
 
+## v0.9.2 - 2023-10-12
+
+This release removes a warning that was being raised when univariate distributions failed to fit and logs the message instead.
+
+### New Features
+
+* When Copulas univariate fit fails, produce a log instead of a warning - Issue [#359](https://github.com/sdv-dev/Copulas/issues/359) by @R-Palazzo
+
+### Maintenance
+
+* Switch default branch from master to main - Issue [#360](https://github.com/sdv-dev/Copulas/issues/360) by @amontanez24
+* Update add-on detection for Copulas - Issue [#362](https://github.com/sdv-dev/Copulas/issues/362) by @pvk-developer
+
 ## v0.9.1 - 2023-08-10
 
 This release fixes problems with the documentation site and drops support for Python 3.7.

{copulas-0.9.2.dev0 → copulas-0.10.0.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: copulas
-Version: 0.9.2.dev0
+Version: 0.10.0.dev0
 Summary: Create tabular synthetic data using copulas-based modeling.
 Home-page: https://github.com/sdv-dev/Copulas
 Author: DataCebo, Inc.
@@ -162,6 +162,19 @@ for specific needs.
 
 # History
 
+## v0.9.2 - 2023-10-12
+
+This release removes a warning that was being raised when univariate distributions failed to fit and logs the message instead.
+
+### New Features
+
+* When Copulas univariate fit fails, produce a log instead of a warning - Issue [#359](https://github.com/sdv-dev/Copulas/issues/359) by @R-Palazzo
+
+### Maintenance
+
+* Switch default branch from master to main - Issue [#360](https://github.com/sdv-dev/Copulas/issues/360) by @amontanez24
+* Update add-on detection for Copulas - Issue [#362](https://github.com/sdv-dev/Copulas/issues/362) by @pvk-developer
+
 ## v0.9.1 - 2023-08-10
 
 This release fixes problems with the documentation site and drops support for Python 3.7.

{copulas-0.9.2.dev0 → copulas-0.10.0.dev0}/copulas/__init__.py

@@ -4,7 +4,7 @@
 
 __author__ = 'DataCebo, Inc.'
 __email__ = 'info@sdv.dev'
-__version__ = '0.9.2.dev0'
+__version__ = '0.10.0.dev0'
 
 import contextlib
 import importlib

copulas-0.10.0.dev0/copulas/visualization.py

@@ -0,0 +1,350 @@
+"""Visualization utilities for the Copulas library."""
+
+import pandas as pd
+import plotly.express as px
+import plotly.figure_factory as ff
+
+
+class PlotConfig:
+    """Custom plot settings for visualizations."""
+
+    DATACEBO_DARK = '#000036'
+    DATACEBO_GREEN = '#01E0C9'
+    BACKGROUND_COLOR = '#F5F5F8'
+    FONT_SIZE = 18
+
+
+def _generate_1d_plot(data, title, labels, colors):
+    """Generate a density plot of an array-like structure.
+
+    Args:
+        data (array-like structure):
+            The data to plot.
+        title (str):
+            The title of the plot.
+        labels (list[str]):
+            The labels of the data.
+        colors (list[str]):
+            The colors of the data.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    fig = ff.create_distplot(
+        hist_data=data,
+        group_labels=labels,
+        show_hist=False,
+        show_rug=False,
+        colors=colors
+    )
+
+    for i, name in enumerate(labels):
+        fig.update_traces(
+            x=fig.data[i].x,
+            hovertemplate=f'<b>{name}</b><br>Frequency: %{{y}}<extra></extra>',
+            selector={'name': name},
+            fill='tozeroy',
+        )
+
+    fig.update_layout(
+        title=title,
+        plot_bgcolor=PlotConfig.BACKGROUND_COLOR,
+        font={'size': PlotConfig.FONT_SIZE},
+        showlegend=True if labels[0] else False,
+        xaxis_title='value',
+        yaxis_title='frequency'
+    )
+
+    return fig
+
+
+def dist_1d(data, title=None, label=None):
+    """Plot the 1 dimensional data.
+
+    Args:
+        data (array_like structure):
+            The table data.
+        title (str):
+            The title of the plot.
+        label (str):
+            The label of the plot.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    if not title:
+        title = 'Data'
+        if isinstance(data, pd.DataFrame):
+            title += f" for column '{data.columns[0]}'"
+        elif isinstance(data, pd.Series) and data.name:
+            title += f" for column '{data.name}'"
+
+    return _generate_1d_plot(
+        data=[data],
+        title=title,
+        labels=[label],
+        colors=[PlotConfig.DATACEBO_DARK]
+    )
+
+
+def compare_1d(real, synth, title=None):
+    """Plot the comparison between real and synthetic data.
+
+    Args:
+        real (array_like):
+            The real data.
+        synth (array_like):
+            The synthetic data.
+        title (str):
+            The title of the plot.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    if not title:
+        title = 'Real vs. Synthetic Data'
+        if isinstance(real, pd.DataFrame):
+            title += f" for column '{real.columns[0]}'"
+        elif isinstance(real, pd.Series) and real.name:
+            title += f" for column '{real.name}'"
+
+    return _generate_1d_plot(
+        data=[real, synth],
+        title=title,
+        labels=['Real', 'Synthetic'],
+        colors=[PlotConfig.DATACEBO_DARK, PlotConfig.DATACEBO_GREEN]
+    )
+
+
+def _generate_scatter_2d_plot(data, columns, color_discrete_map, title):
+    """Generate a scatter plot for a pair of columns.
+
+    Args:
+        data (pandas.DataFrame):
+            The data for the desired column pair containing a
+            ``Data`` column indicating whether it is real or synthetic.
+        columns (list):
+            A list of the columns being plotted.
+        color_discrete_map (dict):
+            A dictionary mapping the values of the ``Data`` column to the colors
+            used to plot them.
+        title (str):
+            The title of the plot.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    if columns:
+        columns.append('Data')
+    else:
+        columns = data.columns
+
+    if len(columns) != 3:  # includes the 'Data' column
+        raise ValueError('Only 2 columns can be plotted')
+
+    fig = px.scatter(
+        data,
+        x=columns[0],
+        y=columns[1],
+        color='Data',
+        color_discrete_map=color_discrete_map,
+        symbol='Data'
+    )
+
+    fig.update_layout(
+        title=title,
+        plot_bgcolor=PlotConfig.BACKGROUND_COLOR,
+        font={'size': PlotConfig.FONT_SIZE},
+        showlegend=False if len(color_discrete_map) == 1 else True,
+    )
+
+    return fig
+
+
+def scatter_2d(data, columns=None, title=None):
+    """Plot 2 dimensional data in a scatter plot.
+
+    Args:
+        data (pandas.DataFrame):
+            The table data.
+        columns (list[string]):
+            The names of the two columns to plot.
+        title (str):
+            The title of the plot.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    data = data.copy()
+    data['Data'] = 'Real'
+
+    if not title:
+        title = 'Data'
+        if columns:
+            title += f" for columns '{columns[0]}' and '{columns[1]}'"
+        elif isinstance(data, pd.DataFrame):
+            title += f" for columns '{data.columns[0]}' and '{data.columns[1]}'"
+
+    return _generate_scatter_2d_plot(
+        data=data,
+        columns=columns,
+        color_discrete_map={'Real': PlotConfig.DATACEBO_DARK},
+        title=title
+    )
+
+
+def compare_2d(real, synth, columns=None, title=None):
+    """Plot the comparison between real and synthetic data for a given column pair.
+
+    Args:
+        real (pandas.DataFrame):
+            The real table data.
+        synth (pandas.Dataframe):
+            The synthetic table data.
+        columns (list[string]):
+            The names of the two columns to plot.
+        title (str):
+            The title of the plot.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    real, synth = real.copy(), synth.copy()
+    real['Data'] = 'Real'
+    synth['Data'] = 'Synthetic'
+    data = pd.concat([real, synth], axis=0, ignore_index=True)
+
+    if not title:
+        title = 'Real vs. Synthetic Data'
+        if columns:
+            title += f" for columns '{columns[0]}' and '{columns[1]}'"
+        elif isinstance(data, pd.DataFrame):
+            title += f" for columns '{data.columns[0]}' and '{data.columns[1]}'"
+
+    return _generate_scatter_2d_plot(
+        data=data,
+        columns=columns,
+        color_discrete_map={
+            'Real': PlotConfig.DATACEBO_DARK,
+            'Synthetic': PlotConfig.DATACEBO_GREEN
+        },
+        title=title
+    )
+
+
+def _generate_scatter_3d_plot(data, columns, color_discrete_map, title):
+    """Generate a scatter plot for column pair plot.
+
+    Args:
+        data (pandas.DataFrame):
+            The data for the desired three columns containing a
+            ``Data`` column that indicates whether it is real or synthetic.
+        columns (list):
+            A list of the columns being plotted.
+        color_discrete_map (dict):
+            A dictionary mapping the values of the ``Data`` column to the colors
+            used to plot them.
+        title (str):
+            The title of the plot.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    if columns:
+        columns.append('Data')
+    else:
+        columns = data.columns
+
+    if len(columns) != 4:  # includes the 'Data' column
+        raise ValueError('Only 3 columns can be plotted')
+
+    fig = px.scatter_3d(
+        data,
+        x=columns[0],
+        y=columns[1],
+        z=columns[2],
+        color='Data',
+        color_discrete_map=color_discrete_map,
+        symbol='Data',
+    )
+
+    fig.update_traces(marker={'size': 5})
+
+    fig.update_layout(
+        title=title,
+        plot_bgcolor=PlotConfig.BACKGROUND_COLOR,
+        font={'size': PlotConfig.FONT_SIZE},
+        showlegend=False if len(color_discrete_map) == 1 else True,
+    )
+
+    return fig
+
+
+def scatter_3d(data, columns=None, title=None):
+    """Plot 3 dimensional data in a scatter plot.
+
+    Args:
+        data (pandas.DataFrame):
+            The table data. Must have at least 3 columns.
+        columns (list[string]):
+            The names of the three columns to plot.
+        title (str):
+            The title of the plot.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    data = data.copy()
+    data['Data'] = 'Real'
+
+    if not title:
+        title = 'Data'
+        if columns:
+            title += f" for columns '{columns[0]}', '{columns[1]}' and '{columns[2]}'"
+        elif isinstance(data, pd.DataFrame):
+            title += \
+                f" for columns '{data.columns[0]}', '{data.columns[1]}' and '{data.columns[2]}'"
+
+    return _generate_scatter_3d_plot(
+        data=data,
+        columns=columns,
+        color_discrete_map={'Real': PlotConfig.DATACEBO_DARK},
+        title=title
+    )
+
+
+def compare_3d(real, synth, columns=None, title=None):
+    """Plot the comparison between real and synthetic data for a given column triplet.
+
+    Args:
+        real (pd.DataFrame):
+            The real data.
+        synth (pd.DataFrame):
+            The synthetic data.
+        columns (list):
+            The name of the columns to plot.
+        title (str):
+            The title of the plot.
+    """
+    real, synth = real.copy(), synth.copy()
+    real['Data'] = 'Real'
+    synth['Data'] = 'Synthetic'
+    data = pd.concat([real, synth], axis=0, ignore_index=True)
+
+    if not title:
+        title = 'Real vs. Synthetic Data'
+        if columns:
+            title += f" for columns '{columns[0]}', '{columns[1]}' and '{columns[2]}'"
+        elif isinstance(data, pd.DataFrame):
+            title += \
+                f" for columns '{data.columns[0]}', '{data.columns[1]}' and '{data.columns[2]}'"
+
+    return _generate_scatter_3d_plot(
+        data=data,
+        columns=columns,
+        color_discrete_map={
+            'Real': PlotConfig.DATACEBO_DARK,
+            'Synthetic': PlotConfig.DATACEBO_GREEN
+        },
+        title=title
+    )

{copulas-0.9.2.dev0 → copulas-0.10.0.dev0}/copulas.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: copulas
-Version: 0.9.2.dev0
+Version: 0.10.0.dev0
 Summary: Create tabular synthetic data using copulas-based modeling.
 Home-page: https://github.com/sdv-dev/Copulas
 Author: DataCebo, Inc.
@@ -162,6 +162,19 @@ for specific needs.
 
 # History
 
+## v0.9.2 - 2023-10-12
+
+This release removes a warning that was being raised when univariate distributions failed to fit and logs the message instead.
+
+### New Features
+
+* When Copulas univariate fit fails, produce a log instead of a warning - Issue [#359](https://github.com/sdv-dev/Copulas/issues/359) by @R-Palazzo
+
+### Maintenance
+
+* Switch default branch from master to main - Issue [#360](https://github.com/sdv-dev/Copulas/issues/360) by @amontanez24
+* Update add-on detection for Copulas - Issue [#362](https://github.com/sdv-dev/Copulas/issues/362) by @pvk-developer
+
 ## v0.9.1 - 2023-08-10
 
 This release fixes problems with the documentation site and drops support for Python 3.7.

{copulas-0.9.2.dev0 → copulas-0.10.0.dev0}/copulas.egg-info/requires.txt

@@ -1,12 +1,11 @@
+plotly<6,>=5.10.0
 
 [:python_version < "3.10"]
-matplotlib<4,>=3.4.0
 numpy<2,>=1.20.0
 pandas>=1.1.3
 scipy<2,>=1.5.4
 
 [:python_version >= "3.10"]
-matplotlib<4,>=3.6.0
 numpy<2,>=1.23.3
 scipy<2,>=1.9.2
 

{copulas-0.9.2.dev0 → copulas-0.10.0.dev0}/setup.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.9.2.dev0
+current_version = 0.10.0.dev0
 commit = True
 tag = True
 parse = (?P<major>\d+)\.(?P<minor>\d+)\.(?P<patch>\d+)(\.(?P<release>[a-z]+)(?P<candidate>\d+))?

{copulas-0.9.2.dev0 → copulas-0.10.0.dev0}/setup.py

@@ -12,13 +12,12 @@ with open('HISTORY.md', encoding='utf-8') as history_file:
     history = history_file.read()
 
 install_requires = [
-    "matplotlib>=3.4.0,<4;python_version<'3.10'",
-    "matplotlib>=3.6.0,<4;python_version>='3.10'",
     "numpy>=1.20.0,<2;python_version<'3.10'",
     "numpy>=1.23.3,<2;python_version>='3.10'",
     "pandas>=1.1.3;python_version<'3.10'",
     "pandas>=1.3.4;python_version>='3.10' and python_version<'3.11'",
     "pandas>=1.5.0;python_version>='3.11'",
+    'plotly>=5.10.0,<6',
     "scipy>=1.5.4,<2;python_version<'3.10'",
     "scipy>=1.9.2,<2;python_version>='3.10'",
 ]
@@ -136,6 +135,6 @@ setup(
     test_suite='tests',
     tests_require=tests_require,
     url='https://github.com/sdv-dev/Copulas',
-    version='0.9.2.dev0',
+    version='0.10.0.dev0',
     zip_safe=False,
 )

copulas-0.9.2.dev0/copulas/visualization.py

@@ -1,164 +0,0 @@
-"""Visualization utilities for the Copulas library."""
-
-import pandas as pd
-
-try:
-    import matplotlib.pyplot as plt
-except RuntimeError as e:
-    if 'Python is not installed as a framework.' in e.message:
-        import matplotlib
-        matplotlib.use('PS')   # Avoid crash on macos
-        import matplotlib.pyplot as plt
-
-
-def scatter_3d(data, columns=None, fig=None, title=None, position=None):
-    """Plot 3 dimensional data in a scatter plot."""
-    fig = fig or plt.figure()
-    position = position or 111
-
-    ax = fig.add_subplot(position, projection='3d')
-    ax.scatter(*(
-        data[column]
-        for column in columns or data.columns
-    ))
-    if title:
-        ax.set_title(title)
-        ax.title.set_position([.5, 1.05])
-
-    return ax
-
-
-def scatter_2d(data, columns=None, fig=None, title=None, position=None):
-    """Plot 2 dimensional data in a scatter plot."""
-    fig = fig or plt.figure()
-    position = position or 111
-
-    ax = fig.add_subplot(position)
-    columns = columns or data.columns
-    if len(columns) != 2:
-        raise ValueError('Only 2 columns can be plotted')
-
-    x, y = columns
-
-    ax.scatter(data[x], data[y])
-    plt.xlabel(x)
-    plt.ylabel(y)
-
-    if title:
-        ax.set_title(title)
-        ax.title.set_position([.5, 1.05])
-
-    return ax
-
-
-def hist_1d(data, fig=None, title=None, position=None, bins=20, label=None):
-    """Plot 1 dimensional data in a histogram."""
-    fig = fig or plt.figure()
-    position = position or 111
-
-    ax = fig.add_subplot(position)
-    ax.hist(data, density=True, bins=bins, alpha=0.8, label=label)
-
-    if label:
-        ax.legend()
-
-    if title:
-        ax.set_title(title)
-        ax.title.set_position([.5, 1.05])
-
-    return ax
-
-
-def side_by_side(plotting_func, arrays):
-    """Plot side-by-side figures.
-
-    Args:
-        plotting_func (callable):
-            A matplotlib function which takes in the standard plot kwargs.
-        arrays (dict[str, np.ndarray]):
-            A mapping from the name of the subplot to the values.
-    """
-    fig = plt.figure(figsize=(10, 4))
-
-    position_base = f'1{len(arrays)}'
-    for index, (title, array) in enumerate(arrays.items()):
-        position = int(position_base + str(index + 1))
-        plotting_func(array, fig=fig, title=title, position=position)
-
-    plt.tight_layout()
-
-
-def compare_3d(real, synth, columns=None, figsize=(10, 4)):
-    """Generate a 3d scatter plot comparing real/synthetic data.
-
-    Args:
-        real (pd.DataFrame):
-            The real data.
-        synth (pd.DataFrame):
-            The synthetic data.
-        columns (list):
-            The name of the columns to plot.
-        figsize:
-            Figure size, passed to matplotlib.
-    """
-    columns = columns or real.columns
-    fig = plt.figure(figsize=figsize)
-
-    scatter_3d(real[columns], fig=fig, title='Real Data', position=121)
-    scatter_3d(synth[columns], fig=fig, title='Synthetic Data', position=122)
-
-    plt.tight_layout()
-
-
-def compare_2d(real, synth, columns=None, figsize=None):
-    """Generate a 2d scatter plot comparing real/synthetic data.
-
-    Args:
-        real (pd.DataFrame):
-            The real data.
-        synth (pd.DataFrame):
-            The synthetic data.
-        columns (list):
-            The name of the columns to plot.
-        figsize:
-            Figure size, passed to matplotlib.
-    """
-    x, y = columns or real.columns
-    ax = real.plot.scatter(x, y, color='blue', alpha=0.5, figsize=figsize)
-    ax = synth.plot.scatter(x, y, ax=ax, color='orange', alpha=0.5, figsize=figsize)
-    ax.legend(['Real', 'Synthetic'])
-
-
-def compare_1d(real, synth, columns=None, figsize=None):
-    """Generate a 1d scatter plot comparing real/synthetic data.
-
-    Args:
-        real (pd.DataFrame):
-            The real data.
-        synth (pd.DataFrame):
-            The synthetic data.
-        columns (list):
-            The name of the columns to plot.
-        figsize:
-            Figure size, passed to matplotlib.
-    """
-    if len(real.shape) == 1:
-        real = pd.DataFrame({'': real})
-        synth = pd.DataFrame({'': synth})
-
-    columns = columns or real.columns
-
-    num_cols = len(columns)
-    fig_cols = min(2, num_cols)
-    fig_rows = (num_cols // fig_cols) + 1
-    prefix = f'{fig_rows}{fig_cols}'
-
-    figsize = figsize or (5 * fig_cols, 3 * fig_rows)
-    fig = plt.figure(figsize=figsize)
-
-    for idx, column in enumerate(columns):
-        position = int(prefix + str(idx + 1))
-        hist_1d(real[column], fig=fig, position=position, title=column, label='Real')
-        hist_1d(synth[column], fig=fig, position=position, title=column, label='Synthetic')
-
-    plt.tight_layout()