Refactor DataVisualization into classes (keeping static API)

[mathparracho]

DataVisualizationcurrently exposes several plotting utilities only as @staticmethods (plot_series, plot_multiple_series, correlation_heatmap, plot_fft, etc.).

The idea is to refactor this so that each plot is implemented as its own class while keeping the current static API working.

This should allow both styles:

# New style (OO)
plot = PlotSeries(data, title="...")
fig, path = plot.plot()

# Old style (still supported)
fig, path = DataVisualization.plot_series(data, title="...")

Proposal

1. Add a common base class:

from abc import ABC, abstractmethod
from matplotlib.figure import Figure

class BaseVisualizer(ABC):
    @abstractmethod
    def plot(self) -> tuple[Figure, str]:
        ...

2. For each existing static method, create a visualizer class:

class PlotSeries(BaseVisualizer):
    def __init__(self, series, title: str = "", save_path: str | None = None):
        self.series = series
        self.title = title
        self.save_path = save_path

    def plot(self) -> tuple[Figure, str]:
        # plotting logic
        path = DataVisualization._save_plot(...)
        return fig, path

3. Keep DataVisualization as a façade that delegates to the classes:

class DataVisualization(ABC):

    @staticmethod
    def plot_series(series, title: str = "", save_path: str | None = None):
        return PlotSeries(series, title=title, save_path=save_path).plot()

    # Other methods follow the same pattern

[thadeuluiz]

Why str output? If needed at all, remember we must be consistent, so at least Paths for outputs.

Furthermore, are we saving on plot?