biophysics_fitting
❭ hay
❭ evaluation
❭ StepThree
StepThree¶
- class biophysics_fitting.hay.evaluation.StepThree¶
Evaluate Step current three.
This class initializes all the evaluation metrics for the
StepThree
stimulus protocol. The empirically observed objectives are in this case:{ "mf3": ["Mean frequency", 22.5, 2.2222], "AI3": ["Adaptation index", 0.0046, 0.0026], "ISIcv3": ["Interspike interval coefficient of variation", 0.0954, 0.014], "DI3": ["Doublet interspike interval", 5.38, 0.83], "TTFS3": ["Time to first spike", 7.25, 1.0], "APh3": ["AP height", 16.4368, 6.9322], "fAHPd3": ["Fast after-hyperpolarization depth", -56.5579, 3.5834], "sAHPd3": ["Slow after-hyperpolarization depth", -59.9923, 3.9247], "sAHPt3": ["Slow after-hyperpolarization time", 0.2131, 0.0368], "APw3": ["AP width", 1.8647, 0.4119] }
See also
_Step
for the template class, andbiophysics_fitting.setup_stim.setp_StepThree()
for more information on the stimulus protocol.- Parameters:¶
soma_thresh (float) – The threshold for APs in the somatic voltage trace. Defaults to \(-30\ mV\).
stim_onset (float) – The onset of the stimulus (\(ms\)). Defaults to \(700\ ms\).
stim_duration (float) – The duration of the stimulus (\(ms\)). Defaults to \(2000\ ms\).
punish (float) – The punishment value in units of \(\sigma\). Used as a baseline if the voltage trace cannot be evaluated on a metric (e.g. if it does not contain an AP). Defaults to \(250 \sigma\).
punish_last_spike_after_deadline (bool) – Whether to punish if the last spike is after the deadline. Defaults to
True
punish_minspikenum (int) – The minimum number of spikes required for this stimulus protocol. Defaults to \(5\).
punish_returning_to_rest_tolerance (float) – The tolerance for returning to rest (\(mV\)). Defaults to \(2 mV\).
definitions (dict) – The definitions for the evaluation metrics. See also:
definitions
.name (str) – The name of the stimulus protocol. Defaults to
'StepTemplate'
.step_index (int) – The index of the step stimulus protocol. Defaults to \(0\). Options are:
[1, 2, 3]
.