Stimulator analysis pipeline

[YudaiSeino]

I wrote stimulator analysis code. Main functions are calc_gain() and calc_timeconstant(). calc_gain() calculates stimulator's signal amplitude. calc_timeconstant() calculates optical time constant and readout delay.

[ykyohei]

It's great to see the simulator pipeline in sotodlib!

Preliminary review comments. sotodlib follows PEP8, but whitespace is missing after commas in many places. I suggest running a linter and correcting them.

[ykyohei]

Thanks! it's getting much better!
I guess I found few bugs, please check.

[ykyohei]

This doesn't necessarily need to be implemented in this PR, but for the site-pipeline of stimulator, we need to implement hk loading from level2 not hkdb.

For example,

sotodlib/sotodlib/site_pipeline/calibration/wiregrid.py

Line 90 in 74fe1ad

def _load_l2_data(config: wg_config, start_time: float, stop_time: float) -> dict:

[ykyohei]

preprocessing look common between gain and timeconstant.
I think it would be cleaner and simpler to make independent preprocessing function.

For example:

def preprocessing(aman, hkdata, idxs=None, n_bins=40, cal_type='all'):

        valid_data = get_encoder_timing(aman, hkdata)  # Get timing against encoder t0
        aman.stm_cal.wrap('sampling_rate', 1 / np.median(np.diff(aman.timestamps)), overwrite=True)
        get_chopping_status(aman)
        get_timing_cut(aman)
        get_signal_temp(aman,hkdata)
        chopping_freqs = <both gain and timeconstant>
        get_coadd_data(aman, "all", n_bins, det_mask)

       # fit_coadd with sine function is also common
        do sine fitting.