ON/OFF dwell-time outputs

write_ONOFFhistograms and write_ONOFFhistograms_key calculate theoretical ON/OFF dwell-time distributions from fitted model rates. The key-based method is recommended for current result folders because it reads model metadata from info_<key>.jld2 rather than inferring the model from a filename.

Key-based use

using StochasticGene

write_ONOFFhistograms_key(
    "results/my-run";
    ratetype = "median",
    bin_size = 100 / 60,
    maxtime = 200.0,
)

StochasticGene rates and times use minutes, so a 100-second imaging interval is 100 / 60 minutes. The call above evaluates the distribution at 100/60, 200/60, ... through 200 minutes.

To use nonuniform or precomputed time points, pass them directly:

write_ONOFFhistograms_key(
    "results/my-run";
    bins = collect(0.5:0.5:100.0),
)

An explicit bins vector takes precedence over bin_size and maxtime.

Output columns

Each ONOFF_<key>.csv contains:

The CDF is computed by the master-equation calculation and the PDF is derived from its finite differences. The CDF columns are therefore preferable when a plot or comparison requires cumulative probabilities; callers no longer need to reconstruct them from the normalized PDF columns.

With simulate=true, the output additionally contains SimON, SimOFF, SimON_CDF, and SimOFF_CDF.

Direct rate-vector use

bins = collect((100 / 60):(100 / 60):200.0)

df = write_ONOFFhistograms(
    rates,
    transitions,
    G,
    R,
    S,
    insertstep,
    bins;
    outfile = "ONOFF_model.csv",
)

The direct method requires a complete model rate vector in the standard rate order. It returns the same DataFrame that it writes.

Folder compatibility method

Legacy rate-file folders can still be processed with:

write_ONOFFhistograms(
    "results/legacy-run";
    bin_size = 100 / 60,
    maxtime = 200.0,
)

For key-based folders, prefer write_ONOFFhistograms_key.