CERAMIC CAPACITOR DC BIAS
========================
Extract all files together. Python 3.9 or newer, standard library only.

    python3 bias_step.py --kind synthetic synthetic-captures.csv
    python3 test_actual_program.py -v

The synthetic dataset was made from independent analytic ramp responses.
The capacitances 9.1, 4.3, 9.8 and 9.8 microfarads are chosen model inputs.

Use observations-blank.csv for paired, time-aligned oscilloscope samples
from one rising edge per run. Save seconds, DRIVE volts and NODE volts.
The four labels are mlcc_low, mlcc_high, film_low, film_high.
For the isolated fixture in the article, replace resistance with the measured
R1 and record the scope input resistance and estimated/measured parasitic.

    python3 bias_step.py --kind observed --resistance-ohm 998.7 \
        --probe-ohm 1000000 --parasitic-pf 150 my-captures.csv

Keep both channels DC coupled and connect all shields to fixture 0 V.
Verify unclipped traces, both plateaus and the intended bias before fitting.
Keep the first 5% and last 10% of samples within their settled plateaus.
The program reports local single-pole estimates and rejects captures outside
the stated screens. This package does not include any instrument traces.
