Currently, color mapping in Tab-right plots uses min/max or assumes values are normalized, which can result in inconsistent or misleading color scales—especially in the presence of outliers. There is a need for a more robust and automatic coloring approach that uses the overall (global) average and standard deviation of the metric values for all plots.

Proposed Solution:

• Remove the requirement for users to select a color scaling method.
• Color mapping will always be inferred automatically using the global mean and standard deviation of the metric values.
• The color scale should cover [mean - kstd, mean + kstd] (with k=2 by default) and values outside this range should be clamped.
• All plots will use this general scaling, ensuring consistent color interpretation.
• Update plot documentation/examples to reflect this automatic, data-driven color mapping.

Benefits:

• Fully automatic and robust color mapping.
• Consistent color interpretation across all plots.
• No need for users to manually select or configure normalization methods.
• More resilient to outliers and varying metric ranges.

Example Implementation:

def normalize_scores(scores, k=2):
    mean = np.mean(scores)
    std = np.std(scores)
    vmin = mean - k * std
    vmax = mean + k * std
    clipped = np.clip(scores, vmin, vmax)
    return (clipped - vmin) / (vmax - vmin + 1e-8)

Additional Context:
This update will simplify plot configuration, improve usability, and ensure that color always reflects the distribution of the data in a statistically robust manner.