edsl 0.1.15__py3-none-any.whl → 0.1.40__py3-none-any.whl

edsl/report/ReportOutputs.py

@@ -1,1228 +0,0 @@
-import platform
-import subprocess
-import tempfile
-
-import copy
-import base64
-import functools
-import inspect
-import markdown2
-import math
-import matplotlib.pyplot as plt
-import numpy as np
-import os
-import pandas as pd
-import seaborn as sns
-import string
-import tempfile
-import textwrap
-import warnings
-import webbrowser
-from abc import ABC, abstractmethod, ABCMeta
-from collections import Counter
-from dataclasses import asdict
-from io import BytesIO
-from IPython.display import display, HTML
-from scipy import stats
-from scipy.stats import chisquare
-from statsmodels.miscmodels.ordinal_model import OrderedModel
-from statsmodels.tools.sm_exceptions import HessianInversionWarning, ConvergenceWarning
-from typing import Callable
-from wordcloud import WordCloud
-from edsl.report.InputOutputDataTypes import (
-    CategoricalData,
-    NumericalData,
-    PlotData,
-    TallyData,
-    CrossTabData,
-    FreeTextData,
-    ChiSquareData,
-    RegressionData,
-)
-from edsl.utilities import is_notebook
-
-
-def save_figure(filename):
-    base, ext = os.path.splitext(filename)
-    if ext.lower() == ".png":
-        plt.savefig(filename, format="png")
-    elif ext.lower() == ".jpg" or ext.lower() == ".jpeg":
-        plt.savefig(filename, format="jpeg")
-    elif ext.lower() == ".svg":
-        plt.savefig(filename, format="svg")
-    else:
-        print("Unsupported file extension. Saving as PNG by default.")
-        plt.savefig(base + ".png", format="png")
-
-
-warnings.filterwarnings(
-    "ignore",
-    category=FutureWarning,
-    module="seaborn._oldcore",
-    message=".*is_categorical_dtype is deprecated.*",
-)
-warnings.filterwarnings("ignore", category=RuntimeWarning, module="statsmodels.*")
-warnings.filterwarnings("ignore", category=RuntimeWarning, module="scipy.optimize.*")
-warnings.filterwarnings("ignore", category=HessianInversionWarning)
-warnings.filterwarnings("ignore", category=ConvergenceWarning)
-
-
-def open_temp_file(file_path):
-    system = platform.system()
-    if system == "Linux":
-        subprocess.run(["xdg-open", file_path])
-    elif system == "Windows":
-        os.startfile(file_path)
-    elif system == "Darwin":  # macOS
-        subprocess.run(["open", file_path])
-    else:
-        print("Unsupported operating system")
-
-
-def convert_svg_to_png_in_memory(svg_bytes):
-    # Create a temporary SVG file
-    with tempfile.NamedTemporaryFile(suffix=".svg", delete=False) as temp_svg:
-        temp_svg.write(svg_bytes)
-        open_temp_file(temp_svg.name)
-
-
-class RegisterElementMeta(ABCMeta):
-    "Metaclass to register output elements in a registry i.e., those that have a parent"
-    _registry = {}  # Initialize the registry as a dictionary
-
-    def __init__(cls, name, bases, dct):
-        super(RegisterElementMeta, cls).__init__(name, bases, dct)
-        if cls.LeftInputType is not None or cls.RightInputType is not None:
-            # Register the class in the registry
-            RegisterElementMeta._registry[name] = cls
-
-    @classmethod
-    def get_registered_classes(cls):
-        return cls._registry
-
-
-def camel_to_snake(name: str) -> str:
-    """Converts a camel case string to snake case, e.g.,
-    >>> camel_to_snake("HelloWorld")
-    'hello_world'
-    """
-    snake_name = ""
-    for index, char in enumerate(name):
-        if char.isupper() and index != 0:
-            snake_name += "_"
-        snake_name += char.lower()
-
-    return snake_name
-
-
-class CustomFunctionWrapper:
-    """A wrapper for a function that adds a name and docstring."""
-
-    def __init__(self, func, name, doc):
-        self._func = func
-        self.name = name
-        self.doc = doc
-
-    def __call__(self, *args, **kwargs):
-        return self._func(*args, **kwargs)
-
-    def __repr__(self):
-        return f"Method: `{self.name}`\nDescription: {self.doc or 'No description available'}"
-
-    def _repr_html_(self):
-        html = markdown2.markdown(
-            f"**Method:** {self.name}\n\n**Description:** {self.doc or 'No description available'}"
-        )
-        # return markdown2.markdown(f"**Method:** {self.name}\n\n**Description:** {self.doc or 'No description available'}")
-        # return f"<b>Method:</b> {self.name}<br><b>Description:</b> {self.doc or 'No description available'}"
-        return html
-
-
-def html_decorator(func: Callable) -> Callable:
-    "A decorator that displays the output of a function as HTML."
-
-    @functools.wraps(func)
-    def wrapper(*args, **kwargs):
-        obj = func(*args, **kwargs)
-        if is_notebook():  # if in a jupyter notebook
-            html = obj.html()
-            return display(HTML(html))
-        else:
-            return obj.view()  # otherwise open in a browser
-
-    return wrapper
-
-
-class Element(ABC, metaclass=RegisterElementMeta):
-    """Base class for all elements.
-
-
-    LeftInputType: The type of the left parent. Could be None.
-    RightInputType: The type of the right parent. Could Be None.
-    OutputDataType: The type of the output data.
-
-    "Root" elements are those that do not have a parent, and are created from the results.
-
-    """
-
-    LeftInputType = None
-    RightInputType = None
-    OutputDataType = None
-
-    def __init__(self, left_parent=None, right_parent=None, output_data=None, **kwargs):
-        self.left_parent = left_parent
-        self.right_parent = right_parent
-        self.left_data = getattr(left_parent, "output_data", None)
-        self.right_data = getattr(right_parent, "output_data", None)
-        self.filename = None
-
-        for key, value in kwargs.items():
-            setattr(self, key, value)
-
-        if (
-            self.LeftInputType is not None
-            and type(self.left_data) != self.LeftInputType
-        ):
-            raise TypeError(f"Left parent must be of type {self.LeftInputType}")
-
-        if (
-            self.RightInputType is not None
-            and type(self.right_data) != self.RightInputType
-        ):
-            raise TypeError(f"Right parent must be of type {self.RightInputType}")
-
-        if output_data is None:
-            self.output_data = self.create_output(
-                self.left_data, self.right_data, **kwargs
-            )
-        else:
-            self.output_data = output_data
-
-    @classmethod
-    def unary(cls):
-        print("Switch to using the cls.element_type method instead")
-        return cls.RightInputType is None
-
-    @property
-    def data(self):
-        print("Shift to using self.output_data")
-        return self.output_data
-
-    @classmethod
-    @property
-    def function_name(cls):
-        return camel_to_snake(cls.__name__)
-
-    @classmethod
-    def element_type(cls):
-        if cls.LeftInputType is None and cls.RightInputType is None:
-            return "root"
-        if cls.LeftInputType is not None and cls.RightInputType is None:
-            return "unary"
-        if cls.LeftInputType is not None and cls.RightInputType is not None:
-            return "binary"
-
-    @classmethod
-    def code_generation(cls, results_name: str, left_column, right_column=None):
-        if cls.element_type() == "unary":
-            return f'{results_name}.{cls.function_name}("{left_column}")'
-        elif cls.element_type() == "binary":
-            return (
-                f'{results_name}.{cls.function_name}("{left_column}", "{right_column}")'
-            )
-        elif cls.element_type() == "root":
-            raise Exception("Should not be called on a root element")
-
-    @abstractmethod
-    def _primary_function(self):
-        "The function that creates the output data, as a dictionary."
-        raise NotImplementedError
-
-    @abstractmethod
-    def _html(self):
-        "The function that creates the HTML representation of the output data"
-        raise NotImplementedError
-
-    def create_output(self, LeftInput, RightInput, **kwargs):
-        if self.element_type() == "unary":
-            output_data = self._primary_function(LeftInput, **kwargs)
-        elif self.element_type() == "binary":
-            output_data = self._primary_function(LeftInput, RightInput, **kwargs)
-        elif self.element_type() == "root":
-            raise Exception("Should not be called on a root element")
-        else:
-            raise Exception("Unknown element type")
-
-        if output_data is None:
-            self.filename = kwargs.get("filename", None)
-            return None
-
-        return self.OutputDataType(**output_data)
-
-    @classmethod
-    def example(cls, **kwargs):
-        class MockParent:
-            def __init__(self, data):
-                self.output_data = data
-
-        left_parent = MockParent(cls.LeftInputType.example())
-        right_parent = (
-            None
-            if cls.RightInputType is None
-            else MockParent(cls.RightInputType.example())
-        )
-
-        return cls(left_parent, right_parent, **kwargs)
-
-    def html(self):
-        return self._html(**asdict(self.output_data))
-
-    def view(self, **kwargs):
-        if hasattr(self.output_data, "buffer"):
-            svg_bytes = self.output_data.buffer.getvalue()
-            convert_svg_to_png_in_memory(svg_bytes)
-        else:
-            if self.filename:
-                print(f"Output was written to file: {self.filename}")
-            else:
-                temporary_directory = tempfile.mkdtemp()
-                with open(os.path.join(temporary_directory, "temp.html"), "w") as f:
-                    f.write(self.html(**kwargs))
-                webbrowser.open(os.path.join(temporary_directory, "temp.html"))
-
-    @classmethod
-    def parameters(cls):
-        return inspect.signature(cls._primary_function).parameters
-
-    @classmethod
-    def create_external_function(cls, results) -> Callable:
-        """Adds a function to the Results class that creates an output element.
-
-
-        In ResultsOutputMixin, there is this function that iterates through the registered
-        classes and adds a function to the Results class for each one.
-
-            def add_output_functions(self) -> None:
-                output_classes = registery.get_registered_classes().values()
-                self.analysis_options = []
-                for output_class in output_classes:
-                    new_function_name = output_class.function_name
-                    new_function = output_class.create_external_function(self)
-                    self.__dict__[new_function_name] = new_function
-
-                    self.analysis_options.append({new_function_name: output_class.__doc__})
-
-        """
-
-        def create_parent(data_type, key, input_type):
-            RootElement = create_root_element(input_type)
-            parent = RootElement.from_results(results, key, input_type)
-            return parent
-
-        if cls.RightInputType is None:
-
-            def func(column, **kwargs):
-                left_parent = create_parent(
-                    *results._parse_column(column), input_type=cls.LeftInputType
-                )
-                return cls(left_parent=left_parent, **kwargs)
-
-        else:
-
-            def func(left_column, right_column, **kwargs):
-                left_parent = create_parent(
-                    *results._parse_column(left_column), cls.LeftInputType
-                )
-                right_parent = create_parent(
-                    *results._parse_column(right_column), cls.RightInputType
-                )
-                return cls(left_parent=left_parent, right_parent=right_parent, **kwargs)
-
-        return CustomFunctionWrapper(
-            html_decorator(func), doc=cls.help(), name=cls.function_name
-        )
-
-    @classmethod
-    def help(cls):
-        help_text = textwrap.dedent(
-            f"""\
-        {cls._primary_function.__doc__}
-        """
-        )
-        # return self._primary_function.__doc__
-        return help_text
-
-
-def create_root_element(output_data_type):
-    class Container(Element):
-        LeftInputType = None
-        RightInputType = None
-        OutputDataType = output_data_type
-
-        def _primary_function(self):
-            raise Exception("Should not be called directly")
-
-        @classmethod
-        def from_results(cls, results, data_name, index=None):
-            data_type, key = results._parse_column(data_name)
-            output_data = results._fetch_element(data_type, key, cls.OutputDataType)
-            return cls(
-                name=data_name,
-                left_parent=None,
-                right_parent=None,
-                output_data=output_data,
-                index=index,
-            )
-
-        def _html(self):
-            return self.output_data.html()
-
-    return Container
-
-
-class PlotMixin:
-    OutputDataType = PlotData
-
-    image_format = "svg"
-
-    @staticmethod
-    def plt_to_buf(plt, format=image_format):
-        buf = BytesIO()
-        plt.savefig(buf, format=format)
-        buf.seek(0)
-        plt.close()
-        return buf
-
-    def _html(
-        self,
-        buffer,
-        title,
-        format=image_format,
-        option_codes=None,
-        width_pct=100,
-        **kwargs,
-    ):
-        image_base64 = base64.b64encode(buffer.getvalue()).decode("utf-8")
-        html = []
-        html.append(title)
-        format_line = "png" if format == "png" else "svg+xml"
-        html.append(
-            f"""<img src="data:image/{format_line};base64,{image_base64}" style="width: {width_pct}%; height: auto;"  />"""
-        )
-        if option_codes is not None:
-            left_option_codes = option_codes.get("left_option_codes", None)
-            if left_option_codes is not None:
-                html.append("<p>Codes:</p>")
-                for key, value in left_option_codes.items():
-                    html.append(f"<p><b>{value}</b>: {key}</p>")
-            right_option_codes = option_codes.get("right_option_codes", None)
-            if right_option_codes is not None:
-                if right_option_codes is not None:
-                    html.append("<p>2nd variable Codes:</p>")
-                    for key, value in right_option_codes.items():
-                        html.append(f"<p><b>{value}</b>: {key}</p>")
-        return "\n".join(html)
-
-
-def tally(responses, options):
-    response_counts = dict(Counter(responses))
-    for key in options:
-        if key not in response_counts:
-            response_counts[key] = 0
-    return response_counts
-
-
-def replace_with_alpha_codes(
-    options: list[str], responses: list[str], prefix: str = ""
-):
-    code_gen = (chr(i) for i in range(65, 91))
-    option_codes = {}
-    for option in options:
-        option_codes[option] = prefix + next(code_gen)
-    new_options = [option_codes[option] for option in options]
-    new_responses = [option_codes[response] for response in responses]
-    return new_options, new_responses, option_codes
-
-
-def header_version(options, index):
-    cleaned_versions = [
-        option.translate(str.maketrans("", "", string.punctuation)).lower()
-        for option in options
-    ]
-    split_versions = [option.split(" ") for option in cleaned_versions]
-    versions = [split_version[:index] for split_version in split_versions]
-    return ["_".join(version) for version in versions]
-
-
-def find_version(options, index):
-    candidate = header_version(options, index)
-    if len(set(candidate)) == len(candidate):
-        return candidate
-    else:
-        return find_version(options, index + 1)
-
-
-def get_option_codes_short_name(options):
-    return dict(zip(options, find_version(options, 1)))
-
-
-def to_strings(split_versions):
-    return ["_".join(version) for version in split_versions]
-
-
-def is_unique(split_versions):
-    return len(set(to_strings(split_versions))) == len(to_strings(split_versions))
-
-
-def get_option_codes(options: list[str]):
-    """Creates a dictionary mapping options to codes."""
-    cleaned_versions = [
-        option.translate(str.maketrans("", "", string.punctuation)).lower()
-        for option in options
-    ]
-    new_cleaned_versions = []
-    shortshands = {"not sure": "not-sure", "need more": "need-more"}
-    for option in cleaned_versions:
-        candidate = copy.copy(option)
-        for key, value in shortshands.items():
-            candidate = candidate.replace(key, value)
-            # print(f"Replacing {key} with {value}")
-            # print(option)
-            # print(option.replace(key, value))
-        new_cleaned_versions.append(candidate)
-
-    cleaned_versions = new_cleaned_versions
-    split_versions = [option.split(" ") for option in cleaned_versions]
-    # get rid of stop words, is possible
-    stop_words = [
-        "a",
-        "an",
-        "am",
-        "the",
-        "of",
-        "and",
-        "or",
-        "to",
-        "for",
-        "in",
-        "on",
-        "im",
-        "that",
-        "are",
-        "i",
-    ]
-    # removes stop words so long as doing so doesn't make responses non-unique
-    for version in split_versions:
-        for stop_word in stop_words:
-            if stop_word in version:
-                index = version.index(stop_word)
-                del version[index]
-                if not is_unique(split_versions):
-                    version.insert(index, stop_word)
-                    # break
-
-    # starts at the end and pops off options if it keeps everything unique
-    # if it goes through and doesn't pop anything for each option, it stops
-    while True:
-        num_pops = 0
-        for version in split_versions:
-            if len(version) > 1:  # we we get to one word, stop
-                removed = version.pop()
-                if is_unique(split_versions):  # no problem
-                    pass
-                else:  # oops, we cut into bone
-                    version.append(removed)
-                    num_pops += 1
-            else:
-                num_pops += 1
-        if num_pops == len(
-            split_versions
-        ):  # stop the loop if we tried popping everyting w/ no luck
-            break
-
-    return dict(zip(options, ["_".join(version) for version in split_versions]))
-
-
-def replace_with_codes(
-    options: list[str], responses: list[str], short_names_dict=None, prefix=""
-):
-    if short_names_dict is not None:
-        option_codes = short_names_dict
-    else:
-        option_codes = get_option_codes(options)
-
-    new_options = [option_codes[option] for option in options]
-    new_responses = [option_codes[response] for response in responses]
-    return new_options, new_responses, option_codes
-
-
-class BarChart(PlotMixin, Element):
-    "Creates a bar chart plot for categorical data."
-    LeftInputType = CategoricalData
-    RightInputType = None
-
-    def _primary_function(
-        self,
-        CategoricalDataObject,
-        width=10,
-        height=5,
-        xlabel="Counts",
-        ylabel="",
-        footer_fontsize=8,
-        title=None,
-        use_code=None,
-        width_pct=100,
-        show_percentage=True,
-        filename=None,
-    ) -> dict:
-        """
-        Generates a bar chart from the provided categorical data object.
-
-        ### Args:
-            - CategoricalDataObject (CategoricalData): An object containing categorical data to be plotted.
-            - `width (int, optional)`: Width of the plot. Defaults to 10.
-            - height (int, optional): Height of the plot. Defaults to 5.
-            - xlabel (str, optional): Label for the x-axis. Defaults to "Counts".
-            - ylabel (str, optional): Label for the y-axis. Defaults to an empty string.
-            - footer_fontsize (int, optional): Font size for the footer text. Defaults to 8.
-            - title (str, optional): Title of the plot. If None, title is taken from CategoricalDataObject.text. Defaults to None.
-            - use_code (bool, optional): Whether to use alphabetical codes for categorical options. Defaults to False.
-
-        Note:
-            If 'use_code' is set to True, each category in the plot is represented by an alphabetical code (A, B, C, ...),
-            and a footer is added to the plot mapping these codes back to the original category names.
-        """
-        responses = CategoricalDataObject.responses
-        options = CategoricalDataObject.options
-        if title is None:
-            title = CategoricalDataObject.text
-
-        option_codes = None
-
-        max_option_length = max([len(option) for option in options])
-        if use_code is None:
-            use_code = max_option_length > 10
-
-        if use_code:
-            if not (d := CategoricalDataObject.short_names_dict) == {}:
-                options, responses, option_codes = replace_with_codes(
-                    options, responses, short_names_dict=d
-                )
-            else:
-                options, responses, option_codes = replace_with_codes(
-                    options, responses
-                )
-
-        response_count = tally(responses, options)
-        total_responses = sum(response_count.values())
-        data = {key: response_count[key] for key in options}
-        data_df = pd.DataFrame(list(data.items()), columns=["Keys", "Counts"])
-        sns.set(style="whitegrid")
-        plt.figure(figsize=(width, height))
-        # sns.barplot(x="Counts", y="Keys", data=data_df, palette="Blues_d")
-        # ax = sns.barplot(x="Counts", y="Keys", data=data_df, palette="Blues_d")
-        ax = sns.barplot(
-            x="Counts",
-            y="Keys",
-            data=data_df,
-            palette="Blues_d",
-            hue="Keys",
-            legend=False,
-        )
-
-        # Adjust layout and add footer if necessary
-        plt.xlabel(xlabel)
-        plt.ylabel(ylabel)
-        plt.title(f"{title}")
-
-        if show_percentage:
-            for p in ax.patches:
-                percentage = f"{100 * p.get_width() / total_responses:.1f}%"  # Calculate percentage
-                x = p.get_x() + p.get_width() + 0.5
-                y = p.get_y() + p.get_height() / 2
-                ax.text(x, y, percentage, ha="center", va="center")
-
-        plt.tight_layout()
-
-        if filename:
-            save_figure(filename)
-            return None
-
-        return {
-            "buffer": self.plt_to_buf(plt),
-            "title": title,
-            "option_codes": {
-                "left_option_codes": option_codes,
-                "right_option_codes": None,
-            },
-            "width_pct": width_pct,
-        }
-
-
-class HistogramPlot(PlotMixin, Element):
-    LeftInputType = NumericalData
-    RightInputType = None
-
-    def _primary_function(
-        self,
-        NumericalDataObject,
-        alpha=0.7,
-        bins=30,
-        xlabel="Value",
-        ylabel="Frequency",
-        color="blue",
-        title=None,
-        max_title_length=40,
-        width_pct=100,
-        filename=None,
-    ):
-        """
-        Generates a histogram plot from a NumericalDataObject.
-
-        This method plots a histogram based on the responses contained within the NumericalDataObject.
-        It allows customization of the plot's appearance including the number of bins, transparency,
-        color, and axis labels. Optionally, a custom title can be set, or it defaults to the 'text'
-        attribute of the NumericalDataObject.
-
-        Parameters:
-            NumericalDataObject (NumericalData): An object containing numerical data and associated responses.
-            alpha (float, optional): The transparency level of the histogram bars. Defaults to 0.7.
-            bins (int, optional): The number of bins in the histogram. Defaults to 30.
-            xlabel (str, optional): Label for the x-axis. Defaults to "Value".
-            ylabel (str, optional): Label for the y-axis. Defaults to "Frequency".
-            color (str, optional): Color of the histogram bars. Defaults to "blue".
-            title (str, optional): Custom title for the histogram. If None, uses the 'text' attribute from NumericalDataObject.
-
-        """
-        responses = [
-            float(x) if x is not None else None for x in NumericalDataObject.responses
-        ]
-        max_title_length = 40
-        if title is None:
-            if len(NumericalDataObject.text) > max_title_length:
-                text = NumericalDataObject.text[:max_title_length] + "..."
-            else:
-                text = NumericalDataObject.text
-        else:
-            text = title
-        plt.hist(responses, bins=bins, alpha=alpha, color=color)
-        plt.title(f"{text}")
-        plt.xlabel(xlabel)
-        plt.ylabel(ylabel)
-        plt.tight_layout()
-
-        if filename is not None:
-            save_figure(filename)
-            return None
-
-        return {
-            "buffer": self.plt_to_buf(plt),
-            "title": text,
-            "option_codes": None,
-            "width_pct": width_pct,
-        }
-
-
-class ScatterPlot(PlotMixin, Element):
-    LeftInputType = NumericalData
-    RightInputType = NumericalData
-
-    def _primary_function(
-        self,
-        LeftNumericalDataObject,
-        RightNumericalDataObject,
-        alpha=0.5,
-        title=None,
-        regression_line=True,
-        x_text=None,
-        y_text=None,
-        width_pct=100,
-        filename=None,
-    ):
-        """
-        Generates a scatter plot using numerical data from two provided data objects.
-
-        This method creates a scatter plot to visually represent the relationship between
-        two sets of numerical data. It offers customization for the plot's transparency
-        (alpha) and title.
-
-        Args:
-            LeftNumericalDataObject (NumericalData): The first numerical data object,
-                                                        used for the x-axis data.
-            RightNumericalDataObject (NumericalData): The second numerical data object,
-                                                        used for the y-axis data.
-            alpha (float, optional): The transparency level of the scatter plot points.
-                                        A value between 0 (transparent) and 1 (opaque).
-                                        Defaults to 0.5.
-            title (str, optional): Title for the scatter plot. If None, a default title
-                                    is generated using the text attributes of the
-                                    NumericalData objects. Defaults to None.
-        """
-        x = LeftNumericalDataObject.responses
-        y = RightNumericalDataObject.responses
-        if x_text is None:
-            x_text = LeftNumericalDataObject.text
-        if y_text is None:
-            y_text = RightNumericalDataObject.text
-
-        if title is None:
-            title = f"{x_text} vs {y_text}"
-
-        plt.title("")
-        plt.xlabel(x_text)
-        plt.ylabel(y_text)
-        plt.scatter(x, y, alpha=alpha)
-
-        if regression_line:
-            slope, intercept, r_value, p_value, std_err = stats.linregress(x, y)
-            x_array = np.array(x)
-            plt.plot(
-                x, intercept + slope * x_array, color="red"
-            )  # Plotting the regression line
-            regression_info_text = (
-                f"Slope: {slope:.3f}\n"
-                f"Std Error in Slope: {std_err:.3f}\n"
-                f"Intercept: {intercept:.2f}\n"
-            )
-            plt.text(
-                0.05,
-                0.95,
-                regression_info_text,
-                transform=plt.gca().transAxes,
-                fontsize=9,
-                verticalalignment="top",
-                bbox=dict(boxstyle="round,pad=0.3", facecolor="white", alpha=0.5),
-            )
-
-        plt.tight_layout()
-
-        if filename is not None:
-            save_figure(filename)
-            return None
-
-        return {
-            "buffer": self.plt_to_buf(plt),
-            "title": "",
-            "option_codes": None,
-            "width_pct": width_pct,
-        }
-
-
-class WordCloudPlot(PlotMixin, Element):
-    LeftInputType = FreeTextData
-    RightInputType = None
-
-    def _primary_function(
-        self,
-        FreeTextDataObject,
-        width=800,
-        height=400,
-        background_color="white",
-        width_pct=100,
-        filename=None,
-    ):
-        """Creates a word cloud plot for free text data.
-
-        Parameters
-        ----------
-        column: str
-            Name of the column in the results to use.
-        width : int
-            Width of the plot in pixels.
-        height : int
-            Height of the plot in pixels.
-        background_color : str
-            Background color of the plot.
-        """
-        responses = " ".join(FreeTextDataObject.responses)
-        text = FreeTextDataObject.text
-
-        wordcloud = WordCloud(
-            width=width, height=height, background_color=background_color
-        ).generate(responses)
-        plt.figure(figsize=(10, 5))
-        plt.imshow(wordcloud, interpolation="bilinear")
-        plt.axis("off")
-        plt.title(f"{text}")
-
-        if filename is not None:
-            save_figure(filename)
-            return None
-
-        #
-        # with open(filename, "w") as f:
-        #    f.write(wordcloud)
-
-        return {
-            "buffer": self.plt_to_buf(plt),
-            "title": "",
-            "option_codes": None,
-            "width_pct": width_pct,
-        }
-
-
-class Tally(Element):
-    LeftInputType = CategoricalData
-    RightInputType = None
-    OutputDataType = TallyData
-
-    def _primary_function(self, CategoricalDataObject, **kwargs):
-        """Creates a tally of responses to a categorical question."""
-        responses = CategoricalDataObject.responses
-        text = CategoricalDataObject.text
-        options = CategoricalDataObject.options
-
-        response_count = dict(Counter(responses))
-        # Add 0s for things that weren't selected even once
-        for key in options:
-            if key not in response_count:
-                response_count[key] = 0
-
-        options.reverse()
-        return {
-            "responses": {key: response_count[key] for key in options},
-            "text": text,
-        }
-
-    def _html(self, responses, text, **kwargs):
-        report_html = [
-            "<div>",
-            f"<p>{text}</p>" "<table>",
-        ]
-        for key, value in responses.items():
-            report_html.append(f"<tr><td>{key}</td><td>{value}</td></tr>")
-        report_html.append("</table>")
-        report_html.append("</div>")
-        return "\n".join(report_html)
-
-
-def compute_cross_tab(left_responses, left_options, right_responses, right_options):
-    left_response_count = dict(Counter(left_responses))
-    right_response_count = dict(Counter(right_responses))
-    # Add 0s for things that weren't selected even once
-    for key in left_options:
-        if key not in left_response_count:
-            left_response_count[key] = 0
-    for key in right_options:
-        if key not in right_response_count:
-            right_response_count[key] = 0
-
-    left_options.reverse()
-    right_options.reverse()
-
-    cross_tab = {
-        left_option: {right_option: 0 for right_option in right_options}
-        for left_option in left_options
-    }
-    # Perform the cross-tabulation
-    for left_response, right_response in zip(left_responses, right_responses):
-        cross_tab[left_response][right_response] += 1
-    return cross_tab
-
-
-class CrossTab(Element):
-    LeftInputType = CategoricalData
-    RightInputType = CategoricalData
-    OutputDataType = CrossTabData
-
-    def _primary_function(
-        self, LeftCategoricalDataObject, RightCategoricalDataObject, **kwargs
-    ):
-        """Creates a cross tabulation of two categorical variables.
-        Parameters
-        ----------
-        left_column: str
-            Name of the column in the results to use for the left side.
-        right_column: str
-            Name of the column in the results to use for the right side.
-        """
-        left_responses = LeftCategoricalDataObject.responses
-        right_responses = RightCategoricalDataObject.responses
-        left_text = LeftCategoricalDataObject.text
-        right_text = RightCategoricalDataObject.text
-        left_options = LeftCategoricalDataObject.options
-        right_options = RightCategoricalDataObject.options
-
-        cross_tab = compute_cross_tab(
-            left_responses, left_options, right_responses, right_options
-        )
-        return {
-            "cross_tab": cross_tab,
-            "left_title": left_text,
-            "right_title": right_text,
-        }
-
-    def _html(self, cross_tab, left_title, right_title, **kwargs):
-        report_html = [
-            "<div>",
-            f"<p>Cross tabulation of: {left_title} and {right_title}</p>",
-            "<table>",
-        ]
-
-        # Assuming all inner dictionaries have the same keys, use the keys from the first one
-        first_key = next(iter(cross_tab))
-        left_options = cross_tab[first_key].keys()
-        headers = [""] + list(left_options)
-        report_html.append(
-            "<tr>" + "".join(f"<th>{header}</th>" for header in headers) + "</tr>"
-        )
-
-        # Fill in the rows of the table
-        for right_option, counts in cross_tab.items():
-            row = [f"<td>{right_option}</td>"]
-            for left_option in left_options:
-                row.append(f"<td>{counts[left_option]}</td>")
-            report_html.append("<tr>" + "".join(row) + "</tr>")
-
-        report_html.append("</table>")
-        report_html.append("</div>")
-        return "\n".join(report_html)
-
-
-class FacetedBarChart(PlotMixin, Element):
-    LeftInputType = CategoricalData
-    RightInputType = CategoricalData
-
-    def _primary_function(
-        self,
-        LeftCategoricalDataObject,
-        RightCategoricalDataObject,
-        num_cols=None,
-        height=5,
-        label_angle=45,
-        title=None,
-        use_code_left=None,
-        use_code_right=None,
-        sharey=True,
-        width_pct=100,
-        filename=None,
-    ):
-        """ "
-            Generates a set of bar plots as a FacetGrid to compare two categorical data sets.
-
-        This method creates a series of bar plots, one for each category in the RightCategoricalDataObject,
-        to compare the frequencies of categories from LeftCategoricalDataObject. The plots are
-        arranged in a grid layout, with an option to specify the number of columns and the height of each plot.
-        Additionally, the angle of the x-axis labels and the title of the grid can be customized.
-
-        Args:
-            LeftCategoricalDataObject (CategoricalData): The first categorical data object,
-                                                        used for the x-axis data in the bar plots.
-            RightCategoricalDataObject (CategoricalData): The second categorical data object,
-                                                        whose categories define the grid columns.
-            num_cols (int, optional): The number of columns in the FacetGrid. If None, it's calculated
-                                    based on the number of categories in RightCategoricalDataObject.
-                                    Defaults to None.
-            height (int, optional): The height of each subplot in the grid. Defaults to 5.
-            label_angle (int, optional): The angle for rotating the x-axis labels for readability.
-                                        Defaults to 45 degrees.
-            title (str, optional): The overall title of the FacetGrid. If None, a default title is
-                                generated based on the texts of the categorical data objects.
-                                Defaults to None.
-            use_code_left (bool, optional): Whether to use alphabetical codes for categorical options
-                                            in the left data object.
-            use_code_right (bool, optional): Whether to use alphabetical codes for categorical options
-            sharey (bool, optional): Whether to share the y-axis across all plots. Defaults to True.
-
-        Notes:
-            - The bar plots are generated using seaborn's barplot function within a FacetGrid.
-            - The layout of the grid is adjusted to accommodate the overall title and to prevent
-            overlap of plot elements.
-        """
-        left_responses = LeftCategoricalDataObject.responses
-        right_responses = RightCategoricalDataObject.responses
-        left_text = LeftCategoricalDataObject.text
-        right_text = RightCategoricalDataObject.text
-        left_options = LeftCategoricalDataObject.options
-        right_options = RightCategoricalDataObject.options
-
-        if use_code_left is None:
-            max_option_length_left = max([len(option) for option in left_options])
-            use_code_left = max_option_length_left > 10
-        if use_code_right is None:
-            max_option_length_right = max([len(option) for option in right_options])
-            use_code_right = max_option_length_right > 10
-
-        if title is None:
-            title = f'"{left_text}" \n by "{right_text}"'
-
-        if len(left_text) > 40:
-            left_text = left_text[:20] + "..."
-        if len(right_text) > 40:
-            right_text = right_text[:20] + "..."
-
-        left_option_codes = None
-        right_option_codes = None
-
-        if use_code_left:
-            left_options, left_responses, left_option_codes = replace_with_codes(
-                left_options, left_responses, prefix="L-"
-            )
-        if use_code_right:
-            right_options, right_responses, right_option_codes = replace_with_codes(
-                right_options, right_responses, prefix="R-"
-            )
-
-        # Figures out how many columns to use in the FacetGrid if not specified
-        if num_cols is None:
-            if len(right_options) < 6:
-                num_cols = len(right_options)
-            else:
-                num_cols = math.ceil(math.sqrt(len(right_options)))
-
-        cross_tab = compute_cross_tab(
-            right_responses, right_options, left_responses, left_options
-        )
-
-        d = {}
-        if use_code_left:
-            d = {v: k for k, v in left_option_codes.items()}
-        left_option_name = d.get(left_text, left_text)
-        if use_code_right:
-            d = {v: k for k, v in right_option_codes.items()}
-        right_option_name = d.get(right_text, right_text)
-
-        df = pd.DataFrame(cross_tab)
-        # Reset index to turn the index into a column
-        df = df.reset_index()
-        # Rename the columns to be more descriptive
-        df.rename(columns={"index": left_option_name}, inplace=True)
-        # Melt the DataFrame to long format
-        df_long = df.melt(
-            id_vars=left_option_name, var_name=right_option_name, value_name="Count"
-        )
-        sns.set(style="whitegrid")
-        # Creating a FacetGrid
-        g = sns.FacetGrid(
-            df_long,
-            col=right_option_name,
-            col_wrap=num_cols,
-            sharey=sharey,
-            height=height,
-        )
-        # ax = sns.barplot(
-        #     x="Counts",
-        #     y="Keys",
-        #     data=data_df,
-        #     palette="Blues_d",
-        #     hue="Keys",
-        #     legend=False,
-        # )
-        # Adding bar plots to the FacetGrid
-        g = g.map(
-            sns.barplot,
-            left_option_name,
-            "Count",
-            order=df_long[left_option_name].unique(),
-            palette="viridis",
-            hue=df_long[left_option_name].unique(),
-            legend=False,
-        )
-        # Rotating x-axis labels for better readability
-        for ax in g.axes.ravel():
-            for label in ax.get_xticklabels():
-                label.set_rotation(label_angle)
-
-        g.fig.suptitle(f"{title}", fontsize=16)
-
-        # Adjust the layout to make room for the title and prevent overlap
-        g.fig.subplots_adjust(top=0.9)  # you can adjust the value as needed
-
-        plt.tight_layout()
-
-        if filename is not None:
-            save_figure(filename)
-            return None
-
-        return {
-            "buffer": self.plt_to_buf(plt),
-            "title": "",
-            "option_codes": {
-                "left_option_codes": left_option_codes,
-                "right_option_codes": right_option_codes,
-            },
-            "width_pct": width_pct,
-        }
-
-
-class ChiSquare(Element):
-    LeftInputType = CategoricalData
-    RightInputType = None
-    OutputDataType = ChiSquareData
-
-    def _primary_function(self, CategoricalDataObject, **kwargs):
-        responses = CategoricalDataObject.responses
-        text = CategoricalDataObject.text
-        options = CategoricalDataObject.options
-
-        response_count = dict(Counter(responses))
-        # Add 0s for things that weren't selected even once
-        for key in options:
-            if key not in response_count:
-                response_count[key] = 0
-
-        observed_counts = list(response_count.values())
-        chi_square, p_value = chisquare(observed_counts)
-        return {"chi_square": chi_square, "p_value": p_value, "text": text}
-
-    def _html(self, chi_square, p_value, text, digits=3, **kwargs):
-        report_html = ["<div>", f"<p>Chi-square test for: {text}</p>" "<table>"]
-        report_html.append(f"<p>Chi-square statistic: {round(chi_square, digits)}</p>")
-        report_html.append(f"<p>p-value: {round(p_value, digits)}</p>")
-        report_html.append("</div>")
-        return "\n".join(report_html)
-
-
-class OrderedLogit(Element):
-    LeftInputType = CategoricalData
-    RightInputType = CategoricalData
-    OutputDataType = RegressionData
-
-    def _primary_function(
-        self, LeftSideCategoricalData, RightSideCategoricalData, **kwargs
-    ):
-        y = LeftSideCategoricalData.responses
-        category_order = LeftSideCategoricalData.options
-        X = RightSideCategoricalData.responses
-        outcome_description = LeftSideCategoricalData.text
-        if not (isinstance(y, list) and isinstance(X, list) and len(y) == len(X)):
-            print(y)
-            print(X)
-            raise ValueError("y and X must be lists of the same length.")
-
-        y_ordered = pd.Categorical(y, categories=category_order, ordered=True)
-
-        # Create a DataFrame from the inputs
-        data = pd.DataFrame({"Outcome": y_ordered, "Predictor": X})
-
-        # Convert the categorical variable into dummy/indicator variables
-        data = pd.get_dummies(data, columns=["Predictor"], drop_first=True)
-
-        for col in data.columns.drop("Outcome"):
-            data[col] = pd.to_numeric(data[col], errors="coerce")
-
-        for col in data.select_dtypes(include=["bool"]).columns:
-            data[col] = data[col].astype(int)
-        try:
-            model = OrderedModel(
-                data["Outcome"], data.drop(columns=["Outcome"]), distr="logit"
-            )  # Use 'logit' for logistic distribution
-            result = model.fit()
-            return {
-                "model_outcome": result.summary().as_html(),
-                "outcome_description": outcome_description,
-            }
-        except Exception as e:
-            return {
-                "model_outcome": f"Error: {e}",
-                "outcome_description": outcome_description,
-            }
-
-    def _html(self, model_outcome: str, outcome_description: str):
-        report_html = [
-            "<h1>Ordered logit</h1>" "<div>",
-            f"<p>Outcome: {outcome_description}</p>",
-        ]
-        report_html.append(model_outcome)
-        report_html.append("</div>")
-        return "\n".join(report_html)