Select subsets to show

[nanel96]

Is there a way to only show certain subsets? The only method I see currently is to set a minimum or maximum size, however this is not very flexible. I like the ability to highlight certain subsets, but it would be more useful for some very large subplots to show only certain subsets of interest. Thanks.

[411an13]

Also looking for an answer to this question! Bumping it.

[411an13]

@nanel96 I made a basic function to select specific subsets to plot. Just feed it the dataframe you'd normally use to make the plot and a list of lists where each sublist corresponds to a subset you want to be included in the plot.

def slice_dataframe_by_index_combinations(df, combinations):
    """
    Slices a MultiIndex DataFrame based on index combinations where the values are `True`.
    All other index levels must have `False` values in the selected rows.

    Parameters:
    - df: A MultiIndex DataFrame.
    - combinations: A list of lists, where each sublist contains the index names 
                    that should be `True` in the desired slice.

    Returns:
    - A subset of the DataFrame where only rows with `True` for the specified index combinations are included.
    """
    # Create a list to hold the conditions for each combination
    conditions = []
    
    # Iterate through the combinations and build a condition for each one
    for combo in combinations:
        # Create a mask for the current combination
        mask = pd.Series(True, index=df.index)
        
        # Go through each level of the MultiIndex and apply the True/False condition
        for level in df.index.names:
            if level in combo:
                mask &= df.index.get_level_values(level) == True
            else:
                mask &= df.index.get_level_values(level) == False

        # Add this condition to the list of conditions
        conditions.append(mask)
    
    # Combine all conditions using the logical OR operator
    final_condition = conditions[0]
    for condition in conditions[1:]:
        final_condition |= condition
    
    # Return the filtered DataFrame
    return df[final_condition]