Add figsize parameter to get_patterns() for customizable plot sizes

[lshpaner]

Changes Summary

1. Added figsize Parameter in get_patterns() Method:

   • Introduced a new parameter figsize to the get_patterns() method, which allows the user to specify the size of the figure when plotting the missing data patterns.
   • Default value is set to (10, 8), which can be overridden by the user.
   • Passed the figsize parameter to the _make_plot() method.

def get_patterns(
    self, 
    X: Matrix, 
    count_or_proportion: str = "count", 
    show_plot: bool = True, 
    figsize: tuple = (10, 8)  # Added figsize as an optional parameter
) -> pd.DataFrame:
    # Other code remains the same
    
    # Pass the figsize to the _make_plot method
    if show_plot:
        self._make_plot(figsize=figsize)

    return self.md_patterns

2. Modified _make_plot() Method to Accept figsize:

   • Updated the _make_plot() method to accept the figsize parameter.
   • This figsize is used when creating the plt.subplots to determine the size of the plot.

def _make_plot(self, figsize: tuple = (10, 8)):
    """"Creates visualization of missing data patterns"""
    
    # Other code remains the same
    
    fig, ax = plt.subplots(figsize=figsize)  # Use the figsize passed as an argument
    
    # Remaining code for plotting
    plt.show()

How It Works Now

• When you call get_patterns() on an instance of mdPatterns, you can pass the figsize parameter to specify the size of the plot.

• The figsize is then used in the _make_plot() method to create the plot with the desired dimensions.

New Implementation

mdp = mdPatterns()
patterns = mdp.get_patterns(X=your_data, figsize=(12, 10))  # Custom figure size

This change provides flexibility to the end user, allowing them to adjust the size of the plot, which helps in preventing issues like scrunched-up labels when dealing with datasets that have many columns or rows.