Use xarray accessor to implement dataset-centric approach

[ldr426]

In order to improve the visualization and interactivity of echogram data, I propose implementing a dataset-centric approach using xarray accessor and leveraging the capabilities of HoloViews. This approach would enable users to manipulate and view echogram data with increased flexibility and customization.

Expected Behavior:

1. Implement an xarray accessor for echogram data, allowing users to access and manipulate the data in a dataset-centric manner.
2. Create methods, like ds.echoshader.view_echogram(), that combines the different components of the echogram visualization (e.g., image, colormap selector, box selection) into a single, customizable layout.
3. Feature like the ds.echoshader.view_echogram() method should accept parameters such as the colormap to be used, with a default value provided (e.g., cmap="jet").
4. Separate out the individual components like follows:

• ds_echogram = ds.echoshader.echogram() - Generate the echogram image component.
• cmapsel = ds.echoshader.colormap_selector() - Provide a colormap selector component for customizing the color map.
• boxsel = ds.echoshader.get_box() - Include a box selection component for highlighting specific areas of interest.

5. Integrate the HoloViews library to enhance the interactive capabilities of the echogram visualization, leveraging features such as linking plots and widgets.

Use xarray accessor: https://docs.xarray.dev/en/stable/internals/extending-xarray.html
https://holoviews.org/user_guide/Linking_Plots.html

[ldr426]

Has been built in #131