Fix fixed-y max calculation to use full plot_data signals

[Copilot]

Fixed-y scaling was deriving y_max from the first non-empty signal in plot_data, which could under-estimate the axis limit when another signal had a higher peak. The fixed-y path now computes y_max from all relevant signals and uses the true maximum.

• Problem summary

  • Fixed-y y-limit selection returned early from _extract_fixed_y_max, so rate_signal could mask a larger participation_signal peak (or vice versa).

• Change made

  • Updated fixed-y max extraction in IPNAnalysis/mea_analysis_routine.py to:
    • iterate both rate_signal and participation_signal
    • collect valid finite positive maxima from each
    • return the global max across collected values (or None if no valid data)

• Effect on behavior

  • Fixed-y plots now normalize against the actual highest plotted network value across stored plot_data signals, improving consistency across wells.

maxima = []
for key in ("rate_signal", "participation_signal"):
    signal = plot_data.get(key)
    ...
    max_val = float(np.nanmax(finite))
    if max_val > 0:
        maxima.append(max_val)

return max(maxima) if maxima else None